The money went physical

The biggest confirmed rounds this week weren’t foundation models or SaaS. They were drones, humanoid robots, satellites and quantum computers.

SVTR’s AI Venture Database logged 86 confirmed rounds (59 disclosed, $7.33B total). Eight of the ten largest are hardware:

• Quantum Systems — $1.2B, military drones (Blackstone & Airbus co-leading)
• Together AI — $800M, AI inference cloud (Aramco, Nvidia, General Catalyst)
• Astribot (智平方) — $735M, humanoid robots
• Chang Guang Satellite — $694M
• Etched — $500M, transformer-only ASIC (Peter Thiel, Jane Street)
• IQM — $234M IPO, quantum computing

Only Together AI and one enterprise-messaging startup are pure software. Robotics, quantum, optics and space fill the rest of the board.

The exits went physical too

Five AI companies IPO’d this week, and 7 of 8 listing / pre-IPO events came from China. Unitree cleared its STAR Market registration in 73 days — the first humanoid / quadruped robot maker to do so on China’s main board. Kuaishou’s Kling closed a ~$3B pre-IPO at a ~$18B post.

This is the closing of a loop. For the first time, China’s AI venture cycle has a working exit anchor, and the scarce resource is shifting from money to a listing path. (We unpack this in this week’s flagship Signal.)

Many-and-small vs few-and-large

A tempting but misleading read is “the top backers are all Chinese.” It’s only half true.

By deal count, Chinese early-stage firms lead — Hillhouse and Sequoia China each backed three rounds. By check size, it flips: the week’s largest rounds were led by Blackstone, Wellington, Nvidia, General Catalyst and Radical. Chinese capital is running a high-volume early-stage factory in hardware; overseas capital is writing a few very large checks. Two different games, same sector.

Why it matters

The model layer is no longer the thing being funded, or the thing being exited. On both the entry and the exit side, capital’s center of gravity moved to things you can touch — robots, chips, satellites, drones.

Full data and the four capital-map charts

The real signal in AI infrastructure this week isn’t another Nvidia challenger. It’s that the largest infrastructure bets all walked around Nvidia’s strongest front.

The two flanks. Qualcomm confirmed a ~$3.9B all-stock acquisition of Modular — a 2022 company that makes no chips. Its product is hardware-agnostic inference software (the Mojo language plus the MAX engine) that lets developers write AI code once and run it across CPUs, GPUs and NPUs without rewriting per chip. A chipmaker that can’t beat Nvidia on raw compute density can still buy the layer that unbinds code from Nvidia. CUDA lock-in is Nvidia’s deepest moat; whoever pries it open makes their own silicon selectable again.

The second flank is light. When single-chip compute is no longer the only scarce thing, how thousands of chips talk to each other becomes the new wall. SpaceX bought one-year-old Mesh Optical (founded by three ex-SpaceX Starlink laser-comms engineers; 1.6Tbps transceivers). MediaTek led HyperLight’s $80M Series C (thin-film lithium niobate photonics). Upscale AI raised a $190M Series A extension for open-standard AI networking — with Nvidia’s own venture arm on the list. None of these fight compute density; they sell connecting, not computing.

Why now. The load shape changed. Training is a peak load that rewards single-chip density — Nvidia’s home turf. Inference is a sustained load that rewards unit cost and the ability to wire huge fleets together efficiently. As the industry shifts from “train a bigger model” to “run models cheaply at scale,” the buyer’s pain moves from “can’t get chips” to two new things: the cost of being locked into one ecosystem, and congestion between chips. When GPU supply was tightest, no one had time for decoupling; as supply eases and inference scales, “don’t get locked in” and “connect the cards well” become fundable businesses.

The China mirror. Where US capital goes around Nvidia, Chinese capital backfills the stack — and the main stage is the secondary market. Enflame won approval for a Shanghai STAR Market IPO; SG Micro (analog) and SinoMicro (lithography gear) completed HK A+H listings on the same day, the latter up 104% on debut. One flanks via M&A and private rounds; the other backfills via public-market funding, one link at a time.

The supply gap. Of the 1,282 founders who’ve applied to SVTR, only ~7% (among those who named a sector) work on compute, chips or semis — the rest cluster in applications. Infra isn’t a founder gold rush; it’s a seller’s market where a few hard teams get chased by capital and acquirers.

Track over the next 90 days: funding and M&A in the decouple-and-runtime layer; quarterly optical-interconnect and silicon-photonics funding; and the share of domestic chip and semi-equipment names in China’s STAR and HK A+H pipeline.

Source: SVTR AI Venture Database, Issue #163 (112 deals).

Everyone is pricing the AI model layer. Goodwater spent a decade pricing the opposite bet.

Goodwater Capital calls itself the world's largest pure consumer-tech VC — ~$3.3B AUM, 79 companies across 19 countries, ~60% in seed and early stage. Its thesis, in co-founder Chi-Hua Chien's words: the biggest winners of the AI era won't sell AI. So it deliberately skips models and compute, and backs consumer applications only.

The historical logic. Chien repeats one rule to LPs: value accrues to the application layer, not infrastructure. Internet era — infrastructure created ~$400B in market cap, applications ~$3.1T. Mobile — infrastructure ~$700B, applications ~$3.7T. Netflix, Meta and Uber all grew on pipes someone else laid.

That distribution is showing up in our own data. Per SVTR's AI venture database, among the AI companies we've tiered, the application layer holds 4,261 companies vs. 571 at the model layer and 671 at infrastructure (as of June 2026). Capital's volume concentrates on that ~10% of model companies; the density of companies actually being built sits in applications. Goodwater is betting that AI won't flip this.

The AI twist: capital efficiency. In 2023 Goodwater raised over $1B in one shot, ~60% for seed and early. The assumption: AI lets application companies hit revenue and margins with tiny teams, so the unit economics of an early check are better than in the last consumer cycle. Chien also leaves infrastructure investors a sharper warning — the capability gap between on-device AI and cloud frontier models was 18–24 months two years ago, is ~6 months now, and he expects it to compress to 3 within a year. Once it closes, the model layer's differentiation commoditizes faster than most expect.

The portfolio: from giant exits to a "personalization" bet. Goodwater's past exits were mostly absorbed by incumbents: Musical.ly (acquired by ByteDance for >$1B in 2017, folded into TikTok), Photomath (Google), Dosh (Cardlytics), Care/of (Bayer). Still-held names span the globe — Monzo in the UK, Toss Bank in Korea. This cycle's AI bets are the part to watch: 22 deals since 2023, almost all seed and early, across entertainment, office, tax, healthcare and finance. Read AI raised twice in 2024 ($21M A, then $50M B six months later); the largest is Stash ($146M H round, 2025). Chien calls the theme "personalization" — entertainment apps doing $100M–$600M ARR at high margins where users don't feel they're using AI; MIDI Health using AI to widen the supply of doctors for perimenopausal hormone therapy. His most contrarian line is "offline reconnection": he backed Fever (a "European Live Nation"), betting that the more infinite digital content becomes, the scarcer and more valuable real-world gatherings get.

The quiet tell. 8 of those 22 deals were co-invested alongside Y Combinator in the same round; Lightspeed and Madrona twice each. So nearly a third of its early bets sit downstream of YC's already-filtered consumer funnel — its real edge is likely less "unique discovery" than judging which app's user retention can survive the giants, and daring to bet early in emerging markets.

The risks — all on one assumption. Consumer-AI moats are thin and get thinner as model capability spills over; differentiation falls back on distribution and brand, the incumbents' home turf. Exits depend on overseas M&A and IPO windows (Getir kept losing money; Monzo had turbulent years), and emerging-market liquidity is more cycle-dependent. And the women's-health / voice / offline-reconnection directions are early signals, not yet proven markets big enough to return a large fund.

Why it matters for cross-border investors. Betting on the application layer is nearly consensus. The hard part is whether the winners are startups at all. When capability is everyone's, the game returns to distribution and user relationships, where Meta and Google already sit. The next 18 months are the test: watch which of Goodwater's bets the giants absorb, and you'll know how deep its moat really is. In China the question barely breathes. Consumer innovation is internalized by ByteDance, Tencent and Alibaba, yet Goodwater is already entangled with Chinese capital (Musical.ly went to ByteDance; portfolio name Weee! serves North American Chinese). Read it as an observation post: watch which country its next seed check lands in.

Sources: SVTR AI venture database (Goodwater record, 22 AI deals, AI company tiering) · Goodwater 2023 fundraise & site · StrictlyVC / TechCrunch interviews with Chi-Hua Chien (June 2026).

Acquirers bought agents, public markets bought scale, and early-stage money bought paradigms with no product yet. The one player caught in the middle — the pure-model company living on narrative premium — is the one getting squeezed.

Buyers paid for agents, not models

Five acquisitions in one week, and not one bought a model. SpaceX took Cursor for ~$60B in all-stock, folding it into the Colossus supercomputer. Salesforce bought Fin (formerly Intercom) for $3.6B to bolt onto Agentforce. Databricks (Panther Labs), 1Password (Apono) and Elastic (DeductiveAI) each bought security or agent capabilities to slot into their platforms. The pricing power has moved: not “whose model is strongest,” but “whose agent already does the work.”

Public markets paid for scale — China edition

The same week, two Chinese AI assets cleared the gate. Xiaohongshu filed in Hong Kong. Enflame, the last of China’s “four AI-chip dragons,” won its STAR Market IPO. Neither is a pure-model company; both are revenue and strategic-position stories. Public markets are pricing cash flow and positioning, not model narrative.

Early stage paid for paradigms

General Intuition raised a ~$300M Series A (reportedly backed by Bezos and Schmidt); CuspAI raised ~$400M for materials-science AI. These are world-model and science bets — funded before there is a product to sell. Capital’s appetite for “a paradigm without a product yet” is unusually high right now.

The middle: OpenAI

OpenAI reported a $38.5B net loss for 2025. Read it carefully: it includes a one-off $41.6B non-cash charge tied to its for-profit transition — strip that out and operations roughly break even. The sharper signal is share: ChatGPT slipped below 50% of the global market for the first time (46.4%), even as it still leads on absolute usage (1.1B MAU).

The honest counterpoint

The single largest confirmed raise of the week was a model company: DeepSeek, at $7.4B — about 74% of all confirmed AI funding we tracked (51 deals, ~$10.05B; 56 companies including 5 rumored rounds). The model layer is not dead. It is bifurcating along US/China lines and concentrating into single points. Geographically the week was a US (23) / China (18) dual core; by count, the application layer dominated (33 companies), while the model layer’s dollars rode almost entirely on DeepSeek.

The squeeze isn’t on “models” broadly. It’s on the middle of the model layer — pure-model companies living on a narrative premium — now pressed from three sides at once: agents above, scale below, paradigm bets ahead.

Capital is pricing the paradigm, not the product — and the US and China are buying two different positions with two different kinds of money.

In SVTR’s AI Venture Database this week (Issue #162, ~56 AI rounds), about ten early-stage deals went to the physical world: world models, embodied AI, robotics, AI materials. Most have little to no revenue. Several were priced at $100M–$2B post-money anyway. The size isn’t the story. The timing is.

1. Born huge: the first real round is already enormous

In a healthy early market, a seed round buys “an idea and a team,” and valuation climbs round by round as milestones are hit. Here, that ladder is gone. One company took a ~$100M “seed++” that is essentially its entire funding to date — it has raised once, and once was nine figures. Another raised a $220M Series A at a $2B post with no prior scaled round.

Valuation has decoupled from milestones. The money doesn’t map to a shipped product; it maps to a founder’s position on a technical paradigm. Capital is making a bet: if this paradigm holds, this team is likely still at the table.

2. Round compression: the confirmation window collapses

The second signal is speed. When the anchor shifts from milestones to position, the cadence shifts from “advance on milestones” to “grab the slot before it closes.” Some companies have run the entire Angel-to-Series B sequence in months; round labels lag far behind the cash already in the door.

The counterintuitive part isn’t “they raised a lot”, it’s the collapse of the confirmation window. The interval between rounds is itself information: time for the company to hit milestones, and a window for the next investor to watch. Run that sequence in months and you’ve given the window up. Investors aren’t adding on more evidence; they’re deciding “wait any longer and I lose the slot.” Round compression is capital hedging missing information with speed.

3. Same paradigm bet, two playbooks

Here’s what only a cross-border view shows. The US and China are buying different positions, with different money.

The US track is software-first, betting on proprietary data flywheels. The checks come from tech billionaires and strategics — Bezos (twice this week), Eric Schmidt, Jeff Dean, Amazon, AMD Ventures, GV, In-Q-Tel. These companies will be priced as “data assets.”

The China track is hardware-first, betting on high-value real-world scenes — hotels and pharmacy retail, e-commerce logistics, 3C and auto manufacturing. The checks come from industrial and manufacturing capital — Wanxiang, BAIC Capital, Didi, Junpu. These will be priced on “scene penetration.”

Three years out, the same “physical AI” label lands in two different valuation models. Benchmark them with one ruler and you’ll be wrong.

Why now

Three forces converged. The LLM dividend window is closing, the strongest models are now free and open-weight, so capital is hunting the next “big and early” paradigm. A cohort of paradigm-grade founders was released from big labs and last-gen unicorns in 2025–2026. And two years of LLM investing proved that hesitating at a paradigm turn costs far more than overpaying one round.

What it means

For founders without a big-lab pedigree: the realistic path is to lock a high-value vertical scene and build data assets others can’t reach, not to fight for the general foundation layer. And remember that a high birth valuation is a liability, not an asset. It raises the bar for every round that follows.

For investors: the funds that judged by model benchmark scores now need a new ruler. Will this paradigm be falsified in three years, is this team truly at its core, and is the path to proprietary data or scenes already locked by an industrial player? In China especially, the best embodied targets may be claimed by industry capital before financial investors even see them.

For major tech companies and platforms: You are both the launchpad for these teams and potential buyers of them in the future. The outflow of top talents will continue, so you need to figure out in advance whether to build internal teams on your own or buy back these startups later.

For enterprise clients: Do not let huge financing rounds sway your purchasing decisions in the short term. Most of these startups are not yet capable of large-scale delivery, and their valuation does not equal their actual delivery capacity.

Tracking over 90 days

Median Round Interval: Tracks the time gap between the current and previous financing rounds for World Model / Physical AI track. If the median shortens to under 4 months, “round compression” is established as a structural trend; if it rebounds above 6 months, the heated financing is more likely an isolated phenomenon rather than a sustained trend.

Proportion of Unprofitable Companies with High Valuations: Calculates the share of Physical AI firms with a valuation of no less than $500 million and no scalable revenue in the current quarter. A rising proportion confirms “anchor shift”, while a falling proportion indicates the market is refocusing on operational milestones.

Sino-US Track Divergence Verification Rate: Monitors counterexamples such as “US enterprises shifting focus to scenario implementation” or “Chinese enterprises prioritizing data flywheel development”. Fewer counterexamples mean the judgment of “divergent development tracks between China and the US” is more credible.

Belief, not fundamentals. NewLimit, co-founded by Coinbase's Brian Armstrong, went from an $810M valuation to $3.1B in just 13 months — entirely before any clinical data. The money holding up that price is a specific kind: crypto and AI exit wealth. Armstrong put in ~$110M himself; the same cluster includes Sam Altman's Retro Biosciences ($180M seed, raising $1B) and Bezos-backed Altos Labs ($3B at launch). They are betting on "reversing aging" as a thesis, not on a de-risked asset.

The science and the moat. NewLimit does partial epigenetic reprogramming: using transcription factors to dial back a cell's aging markers while preserving its mature identity, deliberately avoiding the cancer-prone full Yamanaka route. The real moat isn't a single drug — it's a flywheel. Its Ambrosia AI platform trained on 6.5M T-cell records and finds working factor combinations at over 2x the baseline hit rate; its in-house pxbc-seq cuts screening cost by 100x; mRNA delivery runs about $1.20 a dose. Science is moving fast — its first liver-cell candidate arrived two years ahead of plan — but there is still only one preclinical candidate, and the entire clinic lies ahead.

The mirror: China's pragmatic answer. Per SVTR's AI venture database, life sciences is the most active AI application sector, yet the US dominates 341 deals to China's 15. In the cell-reprogramming cluster, China is essentially one company: Insilico Medicine, which runs AI drug discovery, has 28 pipeline assets (nearly half already in the clinic), and signed a deal worth up to $2.75B with Eli Lilly in March 2026. Tellingly, Lilly is hedging both models — NewLimit's preclinical reprogramming in the US, Insilico's clinical pipeline in China.

Why it matters for cross-border investors. Under the same "longevity" label, the pricing anchor differs entirely: the US bets on a scientific inflection point, with valuation running ahead of data; China ties valuation to clinical progress. A US framework will misprice the Chinese names, and vice versa. NewLimit is best read as a ten-year, high-uncertainty bet on belief — not a company you can value on fundamentals today.

Sources: SVTR AI venture database · Contrary Research · public funding announcements.

For three years, the assumed bottleneck in AI moved in a clean sequence: data, then model capability, then GPUs. SVTR’s latest weekly issue suggests it has moved one layer deeper — into electricity itself.

In SVTR’s AI Venture Database, Issue #161 (80 global AI rounds, June 2026), some of the largest checks went not to model or chip companies, but to companies answering a physical question: where does the power come from, and where does the heat go?

A caveat we hold ourselves to: SVTR’s weekly issue is a curated sample, not a market-wide census. But ~46% of this issue’s rounds (37 of 80) went to physical-world hardtech, and the direction is consistent enough to flag early. The smallest slice — energy, just three deals — is where the founder profile is changing first. That’s the signal.

1. Capital is funding power generation, not smarter models

Three companies that literally generate power raised large rounds in the same window:

SUNUP Fusion — compact controlled fusion on a deuterium–helium-3 fuel cycle; $100M Series A+.

Endurance Energy — deep-sea geothermal tapping heat near plate boundaries; $54M Series A led by Founders Fund.

Weilan Nuenergy Co., Ltd.，small modular reactors (SMRs, pressurized-water design) purpose-built for AI compute; hundreds of millions of RMB across angel+ and Pre-A.

What unites them isn’t a technology roadmap. It’s what they sell: baseload power. Not peak capacity, not a green label — 24/7, predictable, weather-independent electricity. That is precisely what intermittent renewables can’t deliver and what AI data centers most need. When an industry starts paying a premium for baseload rather than peak compute, its binding constraint has shifted a layer.

2. The founders changed: from model researchers to nuclear engineers

If you only look at the dollars, this reads as “energy is hot again.” The deeper signal is in who is starting these companies.

Weilan Nuenergy’s founder spent 20+ years in nuclear engineering, trained at Tsinghua.

SUNUP Fusion’s founder is a Fudan University physics professor with a plasma-physics PhD from UCSD and years running a fusion science institute.

Endurance Energy’s Andrew Redd was a life-support systems engineer at SpaceX, working on Dragon and Starship.

When a field’s top talent migrates from “models and algorithms” toward nuclear engineering, fusion physics and aerospace systems, the genuinely hard, genuinely scarce part of the problem has moved from bits to atoms. Model capability is commoditizing — open weights and APIs turn “smart” into a purchasable input. Getting a reactor onto the grid, or pulling stable geothermal energy from the deep sea, takes a decade of engineering and regulatory relationships. That is the moat software teams can’t replicate on a funding cycle.

3. Power and cooling now define siting — and the US and China diverge

If the scarce input is physical electricity, then where you can get power and where you can dump heat becomes a competitive variable, not a logistics footnote. Two companies in this issue treat siting and cooling as the product itself: Huahong builds fully liquid-cooled edge supercomputing, turning heat dissipation from a data-center constraint into a deployable product; Orbital puts GPUs on satellites, running inference in orbit where power and cooling conditions differ entirely.

The sharper split is in how the two largest markets solve supply. China leans nuclear — fission (Weilan Nuenergy’s SMRs) plus fusion (SUNUP Fusion) — backed by university engineering pipelines and patient state capital. The US is more dispersed — geothermal (Endurance), orbital (Orbital) — backed by dollar funds like Founders Fund and a16z, and talent spilling out of SpaceX and aerospace. For a cross-border investor, “investing in AI’s electricity” is two entirely different underwriting problems.

Why now

The shift is happening now, not earlier, because AI’s load profile changed. Training is a periodic peak load; inference is a continuous baseload. As the industry pivots from training a bigger model to running the model at scale, data-center demand turns from intermittent into a hard 24/7 requirement — and that collides with the grid’s physical limits. A data center takes about two years to build; a new transmission line or a new stable power source takes five to ten. Model gains are slowing and commoditizing while inference demand climbs; the two curves diverging pushed electricity from a background variable to the center of pricing.

Who’s affected

Investors: “powering AI” is becoming its own asset class, but its underwriting looks nothing like SaaS — multi-year timelines, heavy regulation, high capital intensity, exits closer to energy infrastructure than software. Pricing a power plant on software multiples is the easiest mistake for infra funds and software-AI VCs to make. Industrial and deep-tech capital that can read “deliverable, stable megawatts” will capture returns others can’t price.

Founders: at the application layer, inference electricity is becoming a real line in the cost structure, and over time it sets a floor under gross margin. If you’re building infrastructure, the moat is less about technical novelty and more about the ability to secure power — siting, grid connection, long-term power-purchase agreements, regulatory relationships.

Hyperscalers: they’re locking up baseload with long-term PPAs, which both creates demand for independent power and caps it.

Enterprise buyers: remember the long-run conclusion — the floor on inference pricing is set by electricity, not by models.

What we’re tracking (next 90 days)

1. Density and size of energy rounds. Whether new generation deals (nuclear / geothermal / fusion) explicitly tied to AI compute, at $50M+ per round, keep appearing, and whether the median check rises.

2. PPAs and siting. The count of long-term power-purchase or co-location agreements between hyperscale data centers and independent baseload sources (especially nuclear and geothermal).

3. Founder backgrounds. In newly formed “AI infrastructure” companies, the ratio of energy / nuclear / aerospace founders to model-background founders.

SVTR’s AI Venture Database now carries an AI-Energy tag, filter by generation path (fission SMR / fusion / geothermal), cooling and siting (liquid / orbital), and US–China comparison. If you’re evaluating a “power for AI” deal, this taxonomy locates its comparables and where the real risk sits.

On June 5, the AI-tonomy Summit 2026 was held at the headquarters of Plug and Play in Silicon Valley. More than 500 founders, investors and corporate executives attended the event. As a community partner of the summit, SVTR was present throughout the whole event. While most AI conferences over the past two years have focused on competing over AI models, this summit has seen a shift in its agenda focus.

This article will not recap the agenda. Instead, it extracts industrial trends and capital implications observed on-site, tailored for early-stage investors and founders who focus on the commercialization of agents and AI infrastructure.

There were more than 40 speeches throughout the day and 3 parallel sub-forums focusing on enterprise agent implementation, cutting-edge research, and modern AI infrastructure. No one was seriously discussing the countdown to AGI anymore; instead, everyone was talking about one thing: how to truly deploy AI into production systems. To sum up: the first phase of AI competition centers on model capabilities, while the second phase hinges on system implementation.

At the infrastructure roundtable, investors and entrepreneurs including Etienne Dilocker, CTO of Weaviate, Madison Faulkner from NEA, and Kyle McNulty from IQT reached a strong consensus: training remains important, yet the next major battleground lies in inference. Whether enterprises can adopt AI at scale ultimately depends on whether inference costs can be kept affordable, latency can be controlled for production use, and accuracy can meet commercial standards. The decline in token costs has fallen short of expectations, making the inference optimization stack composed of small models, fine-tuning, routing and caching far more crucial than anticipated. Many participants pointed out directly that less than 10% of enterprises have achieved large-scale AI deployment, with major bottlenecks existing in data, access permissions, compliance and governance.

When it comes to value redistribution: model companies captured nearly all the attention in the previous stage, whereas the next wave of dividends will likely go to four key sectors: inference, payment and identity, connectivity and distribution, as well as security.

The Four Layers of Infrastructure

1. Inference Layer: Turning “Adaptive Inference” into a Product

The inference layer has already produced deployable product samples. A representative case is Fastino Labs’ adaptive inference API product Pioneer, which in May 2025 closed a $17.5M seed round led by Khosla Ventures, bringing total funding to $25M. In April 2026, the company released an open-source fine-tuning and deployment package supporting Qwen, Gemma, Llama, and other major open-source models — enabling adaptive inference, automatic model iteration, and live dynamic upgrades. The model a user deploys on day one is the worst version they will ever use.

According to the SVTR database, inference is one of the few infrastructure sectors where independent companies have emerged separately in both the US and China:

• US: Cerebras, Groq, Fireworks, Baseten, Fastino

• China: MOFFETT AI, Enflame, SiliconFlow, QingCheng.ai, and more than ten others working on inference chips or inference stack optimization

This is a track widely favored by cross-border investors.

2. Payment & Identity Layer: Can an Agent Complete a Real Transaction?

The inference layer determines whether an agent can run. The payment and identity layer determines whether an agent can safely complete a transaction — being verified, accountable, and authorized. This is the core bottleneck in landing AI applications: many agents can already compare prices and generate order drafts, yet get stuck at the final payment step.

Two implementation directions have already produced working samples:

Application-layer sample — Snaplii:

• 350,000+ North American users

• 500+ partner brands

• Annual transaction volume exceeding $100M

• 1,700+ agents connected

• Its A2M Skill is natively MCP-compatible and can be directly invoked by large-model clients such as Claude

• It has implemented three principles of agent payment: one-time credentials, tokenization (the agent only sees the token, not the card number), and commercial-grade spend limits and rules that can be revoked at any time

Protocol-layer sample — Verisense:

• Focuses on discovery and interoperability between agents

• Each agent has a verifiable identity

• Behaviors between agents are verifiably recorded into a settlement system

Its development aligns with the latest industry trends. Google's AP2 protocol was released in September 2025 and donated to the FIDO Alliance in April 2026, with the newly added capability of unattended autonomous payment. Ping Identity's Identity for AI became officially available worldwide at the end of March 2026. Protocol-level projects feature long development cycles and rely on cross-ecosystem collaboration. They play a vital role yet have the longest verification periods among all project types.

US-China contrast: Companies in the payment and identity layer are almost entirely concentrated in the US, Europe, and Israel. Independent Chinese startups are largely absent. SVTR judges that domestic agent payments in China will most likely be embedded directly into Alipay and WeChat Pay. The opportunity for Chinese entrepreneurs lies in providing cross-border settlement interlayer services for agents going overseas.

3. Distribution & Security: One is still in the early stage, while the other is quite certain.

1) Distribution layer — still in its early stage

The cognitive sample here is Sense Space, whose thesis is that the “connection effect” of the internet is shifting toward an “awareness effect“ in the AI era:

• In the past, the internet distributed content to people

• In the AI era, capabilities are distributed to agents

This will give rise to new entry points for agent discovery, orchestration, and contextual recommendation. Based on its roadshow materials, this represents a platform revenue potential of over $400 million and an intermediary GMV exceeding $80 billion. However, the prerequisite for realizing this is that the agent population truly explodes.

2) Security layer — demand is already very clear

According to the founder of Horizon3.ai, attacks targeting AI systems have already emerged. Traditional security risks are amplified in the agent era, including:

• Being induced to invoke the wrong tools

• Permission abuse

• Information leakage during inference

We believe that in the coming years, AI-native security will evolve from an auxiliary module into a prerequisite for enterprises to implement AI solutions, and thus become a relatively definite budget priority for corporate procurement.

4. The Hidden Thread and Four Tracking Directions

Beyond the four infrastructure layers lies an industry hidden thread: as the cost of acquiring information through AI keeps falling, what is truly scarce is information credibility — credible people, credible circles, and credible project screening. This is also the core logic behind SVTR’s build-out of a cross-border AI venture capital platform.

In the next phase, SVTR will track four types of companies along these infrastructure layers:

• Inference layer: We focus on unit economics, as demonstration effects are not indicative of actual performance.

• Payment & identity layer: Distinguish two paths with completely different verification rhythms for applications and protocols;

• Connection & distribution layer: Track the real inflection point of explosive agent growth.

• Security layer: Prioritize coverage as a deterministic budget direction.

Those who can build AI with lower-cost inference, safer execution, smoother payment processing and more reliable collaboration hold greater advantages today. The edge of merely pitching AGI concepts is fading away.

From SVTR’s perspective, the most noteworthy takeaway from the latest round of financing is not how many Agent startups have secured funding, but the emergence of an entire cohort of firms dedicated exclusively to Agent governance, whose valuation timelines run well ahead of actual enterprise-level Agent deployments. While the market is still debating how many jobs Agents can replace, investors are already pouring capital into addressing a separate question: once Agents are fully operational, who will monitor their activity, restrict their access permissions, and shut them down in the event of malfunctions.

In short, while the practical capabilities of Agents remain under scrutiny, their inherent risk of going rogue has already been assigned a monetary value. This new intermediate management layer tasked with monitoring, security governance, workflow orchestration and activity logging is evolving into a standalone, revenue-generating infrastructure segment. Its core trait is that it leans more heavily into pure software architecture than execution-layer Agent products, enabling these businesses to command valuation multiples typical of pure software firms. Within the current financing landscape, most general-purpose productivity Agent projects are stuck at small seed-stage funding rounds, whereas several Agent governance startups have hit post-money valuations nearing $200 million at their Series A financing. We previously analyzed vertical Agent tracks in our last briefing—such as Sierra applying software valuation multiples to service-focused business models—and will not revisit that detail here. To draw a clear distinction: our prior coverage focused on which roles Agents displace, while this analysis centers on the third-party entities tasked with overseeing Agents.

Signal 1 | A full cohort of firms focusing on governance has emerged within a single funding batch, while corresponding execution-layer deployment has yet to gain meaningful traction.

First, examine the concentration of these enterprises.

• Gray Swan: Develops adversarial security testing and protection frameworks for models and Agents; secured a $40 million Series A round. Its core team hails from CMU, with founder Zico Kolter, a widely cited scholar in AI safety research.

• Geordie AI: Monitors Agent behaviors and access permissions within enterprise systems; closed a $30 million Series A financing, valuing the company at approximately $180 million post-money, with founding talent sourced from Darktrace.

• Canyon Code: Specializes in orchestrating interactions and dependency relationships between distinct Agents, having raised $5 million in funding.

• Modiqo: Converts validated Agent execution logic into reusable workflow modules, with $3 million in raised capital.

• Trajectory: Conducts ongoing model retraining using real-world end-user interaction data; secured a $15 million seed round resulting in a $115 million post-money valuation, with investment participation from Jeff Dean and Fei-Fei Li.

Adding Tekst, which builds process intelligence solutions, and Tensormesh, focused on cutting inference costs, nearly ten companies in this same funding cohort center their businesses on enabling observability, constraint management and reusability for Agents.

What stands out is this counterintuitive market dynamic: the enterprise-grade Agents these companies are built to support have not yet been widely deployed across industries. The market is building out full-suite management tooling for a technology still in limited mainstream adoption, reversing the typical industry development sequence. Traditionally, a technology scales into widespread use first, operational failures surface thereafter, and governance demands arise reactively in response. By contrast, funding is flowing nearly concurrently to governance-focused and execution-layer Agent developers in this cycle, with financing for certain governance verticals landing even ahead of mainstream Agent rollout.

Here comes the question: What exactly is being purchased with this money if the implementation team has not yet achieved scale growth?

Signal 2 | What enterprises pay upfront to purchase is “controllability” rather than “capability”

The answer lies within clients’ budget structures. The initial budget that enterprises are ready to spend on Agents is shifting focus from “how much work it can accomplish” to “whether it will behave erratically”.

The valuation of Geordie AI best illustrates this trend. Focused on Agent behavior monitoring, permission auditing and risk interception, the firm secured a post-money valuation of approximately 180 million US dollars at its Series A funding round, even though the Agents it monitors have not yet been fully deployed within most enterprises. Investors set such a valuation not based on the value generated by Agents, but on the potential costs stemming from unregulated Agent behaviors. A system capable of stopping Agents from corrupting data, conducting unauthorized operations or leaking confidential information before such incidents occur matters to enterprises not for “efficiency improvement”, but for enabling “confidence in adoption”. Security and observability have evolved into the critical toggle that moves an Agent from PoC to formal production deployment, instead of supplementary fixes implemented post-launch.

Gray Swan presents the same market shift from another perspective. It specializes in pre-launch adversarial testing and red team exercises, turning the verification of “whether an Agent can be manipulated into malicious conduct” into an outsourced service. This business model carries a clear implication: enterprises acknowledge their inability to fully govern Agents in-house and therefore outsource the associated governance risks. This clearly proves that controllability has transitioned from an internal responsibility of engineering teams to an independent procurement category.

Once controllability becomes a standalone paid service, a more fundamental question emerges: given its separability in pricing, does it still follow the same valuation framework as core execution capabilities?

Signal 3｜A divide is emerging between the governance layer and the execution layer, with the former poised to command software-style valuations

This is the most risky yet high-conviction call in this analysis. Valuation rationales for the governance layer and execution layer are diverging.

As pointed out in the previous analysis, the execution layer (vertical Agents in particular) faces an awkward economic model: it replaces professional services, featuring the gross margin structure and delivery costs typical of service-based businesses, yet frequently gets valued as software. The governance layer operates differently. Functions including monitoring, security, orchestration and retraining are inherently horizontal, standardizable and reusable as the number of Agents expands. The marginal cost of onboarding the 100th Agent onto one monitoring system is drastically lower than that of the first, a defining characteristic of genuine software.

This will drive a valuation reversal: market participants assume software valuation multiples belong to task-performing Agents, but in reality, enterprises that oversee Agent operations may be the true beneficiaries of such multiples. Trajectory secured a post-money valuation of $115 million at its seed funding round not by executing specific tasks, but by occupying the horizontal niche of continuous model improvement; every real-world user interaction from new clients fuels and reinforces its unified feedback flywheel. Tensormesh follows a comparable logic: it does not build Agents itself but drives down operational costs for all existing Agents.

This dividing line merits close ongoing monitoring. If this trend holds true, capital markets will increasingly assign software valuation premiums to the horizontal Agent management layer over the near term, while pricing task-focused vertical Agents against metrics aligned with traditional service industries. Put simply, the most software-like segment within the Agent economy may not be the Agent products themselves.

Causes | Why Now

Why is this signal emerging at this moment instead of earlier or later? Three developments have converged within the same timeframe.

First, Agents have evolved from “being able to chat” to “being able to take practical action”, marking a qualitative shift in capability boundaries. Capable of tool invocation, data modification and autonomous sequential decision-making, such Agents now carry concrete and costly risks if out of control, turning governance from an optional add-on into an indispensable prerequisite for deployment.

Second, actual cases of unauthorized operation and malfunctions from leading Agent developers have circulated among enterprises, shifting buyers’ priority from evaluating functional merits to first enquiring about contingency plans for potential failures.

Third, having suffered valuation losses on vertical Agent businesses due to inappropriate software-style pricing for service-oriented offerings, capital investors are proactively hunting for software-like segments across the Agent industrial chain, with governance-related sectors successfully attracting this redirected investment.

Essentially, the current market momentum hinges on investment betting not on Agents’ present capabilities, but their future potential.

Impact | Who Will Be Affected

For entrepreneurs, a clear market positioning window has emerged. If you are building execution-layer Agents, one critical pitfall to watch out for: your core value may not lie in the actual tasks you complete for clients, but in your ability to deliver observable and controllable mechanisms that convince clients to deploy your Agent into production environments. Embedding controllability as an intrinsic product feature rather than an afterthought for compliance directly determines the valuation multiple you can secure in your next financing round. For those developing governance-layer products, sustainable competitive moats do not stem from isolated functional modules—monitoring, testing and orchestration features can all be quickly replicated. Instead, the true edge lies in evolving into a horizontal standard layer that becomes increasingly irreplaceable as more Agents connect to it and more data accumulates; this is exactly the market positioning contested by Trajectory and Geordie.

For investors, this calls for two distinct valuation methodologies. Higher software valuation multiples are justifiable for governance-layer and feedback-layer enterprises thanks to their robust reusability and tangible network effects. In contrast, vertical execution-layer Agent businesses need to be valued strictly based on delivery costs and gross profit margins, and investors should avoid overpaying using software multiples lured by the narrative of AI replacing professional services. A practical due diligence step involves splitting an Agent firm’s revenue into two buckets: income generated from task execution and income from risk assurance services. The proportion of revenue from the latter often serves as a better indicator of the company’s long-term fair valuation than its revenue growth rate.

For large platform operators, foundational model providers will likely integrate core monitoring, logging and access permission functionalities natively into their platforms. Independent governance vendors therefore must pursue deeper technical depth, broader industry coverage and stricter neutrality to sustain viable market space. For enterprise buyers, the primary evaluation benchmark for Agent procurement is shifting from capability demos to security and observability requirements, which will be included in procurement checklists as early as the PoC phase.

Verification: Three Core Metrics Worth Tracking

Over the next 90 days, validate this judgment via three quantifiable metrics.

First, the valuation multiple gap between governance-layer and execution-layer businesses. Track the financing multiples of companies focused on the new round of governance-layer functions (monitoring, security, orchestration and retraining), and calculate their multiple differential against general or vertical execution-layer Agents over the same period. A sustained widening of the gap confirms industry divergence.

Second, the priority ranking of “security and observability” on corporate Agent procurement lists. Based on publicly available enterprise AI procurement cases and RFPs, tally the percentage of instances where “observability, security and compliance” is specified as a mandatory prerequisite instead of an optional add-on. A notable uptick within 90 days validates the second indicator.

Third, the disparity in the number of financing deals for Agent governance-layer ventures between China and the US. Compare newly closed financing rounds targeting Agent governance-layer (excluding execution-layer) businesses in the US, UK and China across the 90-day window. If US-backed deal volume retains its lead with no narrowing gap, the Sino-US structural divergence outlined in the first indicator is substantiated.

Continuing to evaluate agent companies with standard SaaS metrics — LTV/CAC, net retention rate, 120% expansion rate — will systematically underestimate their growth ceiling: global enterprise IT budgets sit at roughly $500 billion; human-related spending (wages, professional services, outsourcing, recruitment costs) is more than ten times that. The underlying AI infrastructure layer is a separate discussion beyond this issue’s scope.

Signal 1 | Professional Domain ARR Growth Already Differs by Order of Magnitude from SaaS

Several data points observable over the past month:

• Contrario, a YC W25 AI recruiting company, reached $6M ARR in under 6 months on a seed round of just $2.3M

• Viktor, an AI coworker embedded in Slack and Microsoft Teams, hit $15M annualized revenue in 10 weeks, raising a $75M Series A

• Searchable, an AI search visibility platform for brands, reached $2M ARR in a few months

• Vétir, an AI outfit selection and wardrobe management app, disclosed 3,500% B2B customer growth and 9x annual revenue

Benchmark: Traditional SaaS companies take a median of 3–5 years to go from zero to $10M ARR. These AI agent companies did it in months.

The difference isn’t product quality — it’s who’s making the buying decision. Contrario’s buyer isn’t the CIO; it’s the HR director. Viktor’s buyer isn’t IT ops; it’s the business team lead. When an HR director is calculating “can this tool save me the cost of a $120K/year recruiting manager,” the decision logic is no longer software procurement — it’s labor substitution. The former has a small budget pool and slow decisions; the latter has a large budget pool and fast decisions.

This leads to the more important question: What exactly are investors pricing in right now?

Signal 2 | Capital Is Systematically Valuing “Professional Services Substitutes”

Reorganizing mid-to-late-stage financings from the past month by the function being replaced:

• Legal: Stilta ($10.5M seed, a16z + YC, AI patent analysis), Lexroom ($50M Series B, European civil law AI law firm operations), LawX ($8.7M seed, AI legal case sourcing system)

• HR / Recruiting: Contrario ($2.3M seed, $6M ARR in 6 months)

• Wealth & Asset Management: Farther ($150M Series D, unicorn status, AUM exceeds $ 23B), Moment ($78M Series C, AUM exceeds $ 10tn)

• Medical Administration & RCM: Commure ($70M raise, $7B valuation)

• Email Security & Enterprise Communications: Ocean ($28M raise, processing over 1 billion emails monthly), Tribal ($10M seed, agents on Salesforce)

These companies boast high valuations when measured by SaaS metrics. However, if we adopt a different benchmark — the total global payroll for the professional roles being replaced — their valuations actually appear conservative. Farther targets the service fee income of global wealth advisors; Lexroom cuts down the billable hours spent on legal research by European law firms; Commure reduces the labor costs of entire teams responsible for revenue cycle management in hospitals.

One detail worth noting: Farther has completed its Series D round, Lexroom its Series B, and Stilta its seed round. Yet their valuation-to-ARR ratios have converged. This indicates that investors believe this track boasts a high success rate and clear exit prospects, so they are willing to offer late-stage valuations for early-stage companies. This represents a capital trajectory entirely different from that of SaaS businesses.

So are the moats of traditional SaaS (product complexity, integration depth, ecosystem dependency) still important? The answer lies in the profiles of the founders.

Signal 3 | Founder Profiles Have Shifted from “SaaS Serial Entrepreneurs” to “Core AI Engineering + Domain Expertise in Parallel”

Representative founder backgrounds from recent financings:

• Hark: Brett Adcock, current CEO of Figure; serial entrepreneur (Vettery acquired by Adecco, Archer went public). New company raised $700M Series A at $6B valuation

• Decart: Dean Leytensdorf, Israeli 8200 military intelligence unit; earned a Technion CS PhD at 23. $300M raised, ~$4B valuation

• Unframe: Shay Levi, co-founder of Noname Security (acquired by Akamai for $450M), second-time founder

• IrisGo: Jeffrey Lai, former Apple Siri engineering manager, incubated by Andrew Ng’s AI Fund

• Modal Labs: Erik Bernhansson, former CTO of Better.com, former engineering manager at Spotify

• Catena Labs: Sean Neville, co-founder of Circle

These founders share one trait: they aren’t people who “know SaaS GTM” — they’re people who “understand core AI capabilities + have judgment in a specific professional domain.” Investors can complete due diligence faster; the resume accounts for half of the process. Early-stage valuations in the agent space are driven up by founder scarcity, not product/revenue metrics. This in itself is a fork in the road from typical SaaS investment logic.

Cause | Why Now

Three variables converged simultaneously from late 2024 to mid-2026, creating the current market:

1. Technology matured. LLMs crossed the reliability threshold for independently completing professional tasks. Repeatable work — legal research, recruiting screening, wealth advisor queries — now has stable product forms.

2. Budget shifted. Beyond enterprise IT budgets, the “human labor budgets” of HR, operations, legal, and finance departments are now formally part of AI procurement discussions. The decision-makers for this budget pool are not in IT — they’re on the front lines of business.

3. Distribution matured. Salesforce, Slack, and Microsoft 365 have opened up agent integration channels. Agent products no longer need to build their own sales and distribution paths — they can embed into existing enterprise entry points. Distribution factors now outweigh product in weighting. For example, Viktor’s ability to reach $15M ARR in 10 weeks owes far more to distribution than to product.

These three things had each happened individually before. Their simultaneous occurrence is what constitutes a brand-new market structure.

Impact | Who Gets Affected

Entrepreneurs: The window for vertical professional domains is shorter than commonly assumed. Don’t build generalized “agent platforms” — beware of being directly displaced. If entering a horizontal track, don’t build a general platform; be the #1 in a vertical domain, owning one function or one job type. The key shift is pricing from “per seat” to “per labor hour / task saved.” Building a general-purpose agent platform has high failure probability, because agents are more scenario-specific and fragmented than SaaS, and the channel economics above them are already forming. One contrarian view: model capability is not the biggest moat — “being able to clearly calculate how many labor hours I save clients each month” is what determines whether you win. Customers will pay a premium for quantifiable human cost savings.

Investors: The valuation anchor needs to change. LTV/CAC still matters, but the ceiling should shift from “this SaaS sub-sector market share” to “total global wage expenditures of the corresponding function.” The core of net retention should shift from “seat expansion rate” to “labor substitution growth rate“ — customer stickiness is no longer a pleasant surprise; what matters is human market share for that function. Agent company return cycles will compress from SaaS’s 7–10 years to 5–7 years, as valuation jumps happen faster and both PE takeover and strategic acquisition exit paths open simultaneously. Early ARR should also be re-framed around “payroll substitution”: customers care more about how much you save them. As long as agent task reliability exceeds what’s promised, retention cycles will be shorter than SaaS but customer expansion will be higher.

Big Platforms: Salesforce, Microsoft, Slack — companies with workflow entry points — are acquiring new strategic leverage. The rules they set for who gets agent channel access will themselves become the source of the next generation of take rates.

Enterprise Buyers: The budget boundary between CIO and CHRO/CFO is beginning to blur. A hybrid procurement committee will emerge in the next 12 months.

Validation: Three Key Metrics to Track

• Within 90 days: Track whether representative companies like Contrario, Viktor, and Stilta continue maintaining ARR growth exceeding 5x annualized rate. Regression to the SaaS median (2–3x) would mean the “payroll substitution” narrative has been over-discounted by the market.

• Within 90 days: Watch for whether any agent company formally writes a “percentage share of saved labor budget” clause into contracts. This is hard evidence that the labor-hour billing model is formally accepted — and will trigger enterprise finance-side accounting classification changes.

• Within 90 days: Watch whether Salesforce or Microsoft launches a clear take-rate mechanism for agent integration (like an App Store model). Once established, channel economics will rapidly reshape early-stage valuations. Horizontal agent companies bound to major platforms will see their valuations surge first.

The timing of the filing was not coincidental: on May 18, a California federal jury dismissed all of Elon Musk’s claims against OpenAI, clearing the most critical legal obstacle in the IPO process. In September, CFO Sarah Friar favored a longer preparation timeline. The current filing reflects CEO Sam Altman’s push for a September listing, with the CEO’s position prevailing over internal disagreements on pacing.

The article argues that the real signal of this filing is not the label of “world’s largest AI IPO”, but rather that OpenAI chose to proactively lock in public market pricing power and narrative control at the moment when the three battlegrounds — ARR growth rate, operational profitability, and valuation multiples — are about to fully diverge from Anthropic. This is a defensive preemption, not a winner’s posture.

I. ARR Growth and Valuation Multiples Have Already Reversed

When the annualized revenue trajectories of both leading companies are plotted together, the reversal inflection point appears in Q1 2026:

• OpenAI rose from $20 billion ARR at end-2025 to approximately $24 billion by April — ~25% growth in 5 months;

• Anthropic surged from $9 billion to $30 billion (official disclosure) over the same period, and according to follow-up reports from The Information and SemiAnalysis, the run rate was approaching $44 billion by May.

• OpenAI Q1 actual revenue was $5.7 billion; Anthropic Q1 was $4.7 billion — the gap narrowing to a $1 billion range.

Valuation multiples calculated on current disclosed bases:

• OpenAI: $852B / $24B ARR ≈ 35x

• Anthropic: $380B / $14B ARR ≈ 27x in February funding round negotiations; $900B / $44B ARR ≈ 20.5x in this month’s new negotiations

The implied secondary market pricing for Anthropic has also touched $1 trillion. As Anthropic overtakes on ARR, valuation multiples are converging — meaning OpenAI can no longer sustain a 35–40x ARR multiple on the basis of “industry leader” scarcity premium. This is the direct trigger for the preemptive filing.

II. Is the $1 Trillion Anchor Sustainable?

Bull case for $1 trillion target valuation has three pillars:

1. Private valuation trajectory rising consistently for 12 months: ~$157B at end-2024 → ~$500B in October 2025 → $852B this round;

2. Microsoft remains both largest customer and largest compute provider — the partnership framework provides a capital markets narrative of locked-in revenue streams;

3. Dual consumer and enterprise lines — business structure is broader than API-only Anthropic.

Bear case data is equally clear:

1. Absolute growth has been overtaken: Anthropic grew ARR 4.9x in 5 months; OpenAI only 1.25x over the same period. Even at a 40x ARR target, OpenAI’s pricing is less justified than Anthropic’s current implied 20–22x multiple;

2. Operational leverage is moving in opposite directions: Anthropic may achieve its first operational profit in Q2; OpenAI’s losses are widening over the same period. CFO Friar — bound by a 5-year, $600 billion compute commitment — has publicly acknowledged that revenue growth is insufficient to support all data center construction. This is an unavoidable risk exposure in the S-1 roadshow;

3. Governance terms unresolved + banker conflicts: Revenue sharing with Microsoft, IP ownership, and board power dynamics are not fully publicly disclosed. The underwriting syndicate of Goldman Sachs and Morgan Stanley are simultaneously in talks with Anthropic about its IPO. The same roadshow team must pitch two direct competitors to the same pool of LPs — this is yet another motivation for OpenAI to file first.

III. Transmission to the financing environment of the track

OpenAI’s filing will squeeze competitors on two levels:

1. Infrastructure comparable anchor reset

GPU cloud and training platforms like CoreWeave, Lambda Labs, and Together AI — once detailed OpenAI compute expenditure data becomes available as a benchmark — will face recalibration of their current 9–15x revenue multiples in private market pricing. Some companies may be forced to prematurely disclose take-or-pay contract details.

2. Window compression effect

SpaceX publicly filed its S-1 on May 20 (target valuation $1.75 trillion, raising $75 billion, Nasdaq ticker SPCX), with a roadshow scheduled to launch June 5 and pricing around June 12. OpenAI confirmed its filing just two days after SpaceX’s S-1 went public — the time-window pressure is immediate. Q3–Q4 will see at least $150 billion combined in mega-issuances from SpaceX and OpenAI. The listing window for mid-tier AI companies is effectively pushed to Q1 2027 or later.

For Anthropic, however, OpenAI’s filing timing actually presents a tactical opportunity: its proposed $900 billion private round can be completed before OpenAI’s public S-1, positioning Anthropic as the “unlisted but ARR- and profit-stronger” comparable before its own roadshow.

IV. Three Variables to Watch Over the Next 1–2 Quarters

1. OpenAI’s public S-1 disclosure — gross margin and Microsoft revenue-sharing mechanism

The public S-1 is expected 15 days after the confidential filing, i.e., no earlier than mid-to-late August. Key metrics to watch: if gross margin is significantly below secondary market implied assumptions, or if the Microsoft revenue-sharing mechanism is above market-estimated take rates (~20% range), the $1 trillion valuation will likely fall back to the $750B–$850B range.

2. Whether Anthropic files before OpenAI’s public S-1, and its disclosed Q2 operating profit data

If Anthropic files and can demonstrate Q2 profitability, OpenAI will face the “industry leader” narrative being dismantled. OpenAI’s valuation would face substantive downside, with the lower bound potentially breaking through $700 billion.

3. SpaceX’s June offering results and final pricing range

SpaceX’s current target is $1.75 trillion at $75 billion raise. If the final issuance valuation falls below $1.5 trillion, or the price drops to the lower end of the issuance range within 30 days after pricing, it indicates that the market’s capacity to absorb trillion-dollar landmark IPOs is nearly saturated. This will directly affect the difficulty of securing anchor investors and the pricing ceiling for OpenAI’s roadshow in September.

On May 18th, SVTR partnered with Zhiwai to host a closed-door luncheon on “Chinese AI Brands Going Global” in Silicon Valley. The event gathered entrepreneurs building consumer brands, AI applications, cloud services, influencer marketing, e-commerce platforms, and deep-tech startups on US marketplaces.

Participant niches included: consumer brands for specific US demographics, AI-powered influencer marketing, UGC licensing, content platforms for micro-merchants, deep-tech ventures, and cloud infrastructure.

The core question everyone grappled with: As AI, supply chains, content channels, and global markets shift simultaneously, how should Chinese entrepreneurs rethink “brand globalization”?

Deep Needs Behind a Small Niche

After the event, Zhiwai founder Lina shared a representative case: An entrepreneur building an electric vehicle for the American hunting community.

At first glance this seems extremely niche. But digging deeper, the needs run surprisingly deep:

The US has a large population of licensed hunters who are highly willing to invest in gear. But standard e-bikes completely fail their real scenarios: climbing hills with gear, navigating muddy terrain, cold-weather starting, long-range needs, chain-break resistance, and high climbing ability.

What hunters actually need: a vehicle that handles 30-degree slopes, starts at -20°F, carries 200 lbs of gear, and lasts a full day on a single charge.

Even more noteworthy: the entrepreneur himself is the target user. The project started from “I’m fed up with this problem myself” — not “I’m going to build something for a big market.”

This case sparked the core discussion: Must great entrepreneurial opportunities start from “big markets”? The answer is no.

Often, a small circle actually has sharper pain points, more genuine demand, and stronger word-of-mouth — making it easier to carve out a true niche market. Small community, high pain point, strong reputation — this may well be the starting point for a global brand.

Especially in the US market, segmented user groups have very specific needs and clear willingness to pay: hunters, outdoor enthusiasts, micro-merchants, content creators, or professional users — they don’t need “everything to everyone.” They just need a tool that truly understands their scenario and solves their specific problems.

This is what entrepreneurs need to rethink: Going global isn’t just exporting domestic goods; it’s about using global supply chains, AI tools, and local user insights to redefine the product entirely.

AI’s Penetration into Brand Globalization

The session also covered AI’s real-world applications in traditional industries: Even in manufacturing like electric bicycles, AI has entered the product development process. The hunter e-bike startup is already using LLMs to assist with code writing, drivetrain design, and generating product mockups.

AI’s impact extends far beyond software and marketing — it has seeped into manufacturing and hardware entrepreneurship.

In the past, building a hardware brand required a long cycle: product definition, engineering design, supply chain coordination, market testing, and content distribution. Now AI helps teams move faster on solution design, prototype expression, content production, user feedback analysis, and market testing — dramatically improving efficiency.

But AI doesn’t replace an entrepreneur’s understanding of users. It amplifies the advantages of teams that truly understand users, scenarios, and supply chains.

Four Typical Paths for Chinese AI Brand Globalization

1. Leveraging China’s supply chain to serve niche US markets
Like the hunter e-bike — the essence isn’t “e-bike exports,” but redefining products around high-pain-point user groups.

2. Next-generation brand infrastructure powered by AI
Influencer marketing, creator discovery, video content analysis, and UGC licensing. Projects like Influcio, Chimeboard, and Vyrill help businesses better understand content, influencers, users, and conversion logic.

3. New store-opening platforms for micro-merchants
Aladdinmall focuses on online commerce, local discovery, pre-purchase communication, managed payments, and mobile shopping — lowering barriers for local businesses to go online.

4. Deep-tech ventures from foundational technology
Breaking through from the technology side to find industrial application scenarios for global markets.

AWS and startup solutions architects also contributed perspectives on cloud infrastructure, MVP architecture, AI agents, LLM application development, and modern data stacks — supporting startups as they scale from early product to growth.

All of this points to one trend: Brand globalization is no longer “selling Chinese goods overseas.” It has become a system-wide endeavor: product definition, supply chain organization, user insights, content marketing, channel building, AI tools, data infrastructure, and local trust networks.

For Chinese entrepreneurs, the US market is both a proving ground and an amplifier: real user needs, diverse niche markets, mature commercial infrastructure, and customers willing to pay for professional products. But it also demands that entrepreneurs truly understand local users — rather than equating “globalization” with cross-border sales.

Great brands often don’t start from a grand slogan, but from a real user’s real pain point.
A small niche may hide a deep need.
A specific user group may lead to a global brand.

This is the direction SVTR aims to connect: gathering practitioners in products, brands, technology, supply chains, traffic, and industrial services — letting real experience flow and helping more Chinese entrepreneurs discover real opportunities in global markets.

SVTR believes this is one of the most underrated secondary signals of 2026. A group of elites with the most solid information advantages are voting with their feet to redefine the real moat of the vertical Agent track. It lies neither in models nor data, but in category insights and go-to-market strategies. The homogenization at the model level has driven the marginal value of merely understanding technology down to nearly zero. Meanwhile, the genuine barriers of vertical Agents happen to fall within the core strengths of these VC partners.

It is noteworthy that such career shifts have not taken place in seemingly grander fields such as model development, foundational AI enterprises or chip sectors. Even the most high-profile tracks have failed to attract partners to step into actual business operations. As the window for information arbitrage on deal sourcing narrows, they opt to participate in the industry in person rather than making more aggressive investments.

Signal 1 | Partner migration is highly concentrated in the "enterprise back-end"

Over the past 12 months, there have been at least seven clearly documented cases where partners from mainstream US dollar funds left their positions to start their own businesses:

• a16z GP Anjney Midha → Amp: AI compute tuning, first-round raise exceeding 130M

• Kleiner Perkins GP Joublin Mirzadegan → Roadrunner: AI-native CPQ, with a total funding of 27 million US dollars raised through seed and Series A funding

• a16z Partner Jack Soslow → Ciridae: an AI operation system for mid-sized enterprises, which secured $20 million in seed funding

• Founders Fund Partner Sam Blond → Monaco: an AI sales platform, which raised $50 million in Series B funding with a valuation exceeding $500 million

• Basis Set Venture Partner Ian Crosby → Synthetic: an AI accounting firm, which obtained $10 million in seed funding

• Rough Draft VP Sahir Jaggi → Forus: an AI prescription and insurance process platform, which closed a $160 million initial round and became a unicorn enterprise

• Twitchex CSO, angel investor with investments in over 50 companies ColinCarrier → OriginLab: a provider of world model training data, which raised $8 million in seed funding.

If we broaden the scope to include senior investors and serial entrepreneurs, and further add Cowboy Space co-founded by Baiju Bhatt, co-founder of Robinhood (raised $275 million in Series B funding with a valuation of $2 billion), Exponent founded by Charlie Cheever, co-founder of Quora, and Recursive Superintelligence led by Richard Socher, former Chief Scientist of Salesforce (post-money valuation of $4.65 billion), the figures will rise even higher.

What makes this list counterintuitive is not the fact that these people have left their previous positions, but that they have collectively steered clear of the most high-profile tracks such as large models, chips and consumer-oriented AI.

Signal 2 | Not Betting on Technology, But on “GTM × Category Insight”

Analyzing what these companies are offering: CPQ solutions, accounting bookkeeping, sales compensation management, prescription pricing, franchisee financial management, insurance brokerage services, sales process management and more. Almost all of them are enterprise back-end businesses that are mundane yet deeply accumulated, entail high switching costs and require mastery of industry-specific terminology. The real barriers to entry in these sectors have never lain in model selection, but in the tripartite combination of industry expertise, client relationships and sales rhythm control.

What core strengths do VC partners possess? They have built up insights into customer profiling, sales rhythm mastery and valuation benchmarking by evaluating 2,000 companies over a decade. These are precisely the most scarce resources for vertical Agent companies in their rapid expansion phase. The biggest bottleneck for founders of vertical Agent enterprises is usually not how to train Agents, but how to get involved in procurement negotiations, set reasonable pricing benchmarks and secure the first batch of reference clients. Former general partners hold inherent structural advantages in these three aspects.

In other words, they did not enter the industry merely drawn by the AI boom; instead, they realized their inherent strengths could for the first time be monetized on the seller side. If such core strengths can be revalued, what changes will take place across the entire VC value chain?

Signal 3 | The Sourcing Era Is Ending; The Underwriting Era Is Just Beginning

This is truly a third-tier pivotal signal. If the optimal strategy of frontline GPs has shifted from “finding the next Forus” to “establishing the next Forus independently”, the value chain of the VC industry is being rewritten.

Over the past five years, excess returns from AI investments mainly stemmed from sourcing advantages — being the first to spot OpenAI, Anthropic, or issue term sheets to Cursor. However, vertical Agent projects are overly saturated and fragmented, and the technical gaps are too narrow for GPs to maintain prominent sourcing edges. The likelihood of five high-caliber AI Sales Agent startup teams emerging simultaneously is far higher than that of five equally strong foundational AI company teams.

As information arbitrage opportunities in sourcing fade away, excess returns will flow back to two types of participants: funds equipped with industrial know-how to conduct professional underwriting, and partners with GTM leverage capable of launching startups on their own. The logical closed loop behind this trend is fully formed. The only remaining question is: why did this trend break out intensively between 2025 and 2026, rather than in 2023 or 2024?

Causes | Why Now?

Three major forces are taking effect simultaneously:

1. Technological inflection point. The marginal capability gaps at the model level are narrowing to zero, forcing vertical Agents to shift their competitive moat toward go-to-market strategies and category insights. In 2024, startups could still secure Series A funding by claiming “superior Agent orchestration technology”, yet this growth narrative has lost its appeal by 2026.

2. LP inflection point. Starting from Q3 2025, dollar limited partners have grown far less patient with reinvesting in foundational AI companies, and capital is flowing toward application-layer ventures with proven revenue streams and verifiable unit economics. Leading general partners can directly sense this shift in sentiment during LPAC meetings, at least six months earlier than such trends appear in secondary research reports.

3. Incentive inflection point. The carry profit realization timeline of top-tier GPs is no longer aligned with the market opportunity windows they identify. If a GP forecasts a seed-stage project they would traditionally back could surge tenfold in value within five years, while carry profits locked in the fund take eight years to vest and get split among a 20-person core team, the expected returns from personal direct investments outpace long-term career gains within institutional funds.

This checklist comes into being precisely as these three inflection points converge.

Impact | Who Gets Hit

Founders: The profile of real competitors has shifted. Over the past 18 months, vertical Agent entrepreneurs have mainly faced competition from peer founders and incumbent software companies. Going forward, this competitive landscape will welcome a new group of “ex-GP founders” who come equipped with limited partners (LPs), go-to-market (GTM) networks, client rosters and secondary valuation benchmarks. Their edge lies not in products, but in pre-established financing rhythms and sales networks. They can mobilize over 20 million US dollars at the seed round stage, and skip traditional demos to directly connect with strategic LPs in Series A financing. Ordinary founders need to rethink their differentiated positioning: superior technology is no longer sufficient. They must prove they possess irreplicable industrial expertise in specific niche sectors that even GPs cannot match, such as exclusive client connections, unique data flywheels, or hard-to-replicate founder-market fit within the industry. A concrete strategic takeaway is that instead of pursuing general horizontal Agents, it is wiser to firmly anchor the business in vertical industries overlooked by GPs.

Investors: The era of deal-sourcing arbitrage is drawing to an end. A clear inference is that if even firm partners are launching their own startups, the scarcity of merely discovering investment deals will inevitably decline, while the scarcity of accurately evaluating projects will rise sharply. The moat of early-stage funds is shifting from gaining early access to projects to identifying industry ceiling potentials ahead of others. This means capabilities once regarded as cost centers, including investment research teams, industry analysts and exit return models, are becoming core factors in investment valuation. Meanwhile, the transferable value of secondary valuation benchmarks and LP connections is growing increasingly prominent, and early-stage personal connections alone can no longer sustain the differentiated competitiveness of new funds. Mid-sized funds need to stay highly vigilant; under this new supply-side logic, funds that are neither large enough to provide underwriting support nor close enough to participate directly in industrial layout are in the most precarious position.

Platforms and Major Enterprises: These vertical Agent enterprises pose the most direct substitution threat to SaaS giants including Salesforce, ServiceNow and Workday within the next 24 months, and merger and acquisition activities are expected to surge notably starting from the second half of 2026.

Enterprise Clients: CIOs and CFOs are no longer faced with the simple choice of whether to adopt AI tools. Instead, they have to decide between purchasing AI modules developed by established incumbents or brand-new AI-native platforms built from scratch. The latter have already accumulated client cases with annual recurring revenue (ARR) ranging from 600,000 to 1 million US dollars.

Verification: Three Key Indicators Worth Tracking

1. Within 90 days, check if more than two current partners from a16z, Founders Fund, Kleiner Perkins and Sequoia resign to start their own businesses, with their business focus still centered on enterprise backend Agents. If this happens, it indicates a structural industry shift rather than isolated individual cases.

2. Verify whether the median post-money valuation of seed-stage vertical Agent enterprises exceeds 50 million US dollars before Q3 2026. This is a core threshold to judge whether general partners are systematically lifting the bottom valuation of this track.

3. Confirm if there emerges the first public case of seed financing initiated by former GPs with participation from their former venture capital funds within 90 days. This will officially signal the recognition of this new supply-side logic by limited partner capital.

It’s a beautiful line. It casts the investor as a talent scout with preternatural instincts, places the founder at the center of the story, and wraps every check in a narrative about betting on human potential. It’s gospel in Silicon Valley. It’s even more so in the official talking points of China’s top-tier VC firms. But here’s what we believe at SVTR: in today’s AI investing landscape, this phrase has degraded from a philosophy of judgment into the single biggest fig leaf covering the industry’s most systematic failures — and the bill is coming due at an alarming pace.

More precisely, “backing people, not ideas” is no longer a form of judgment. It’s a way of avoiding judgment. When an investor says this, they almost never have to explain what they actually saw. The phrase has three convenient properties: it conforms to industry norms, it doesn’t offend founders, and it requires zero specific argumentation. Naturally, it becomes the perfect cover for every bias, every résumé worship, every shortcut.

A Philosophy Misread for Four Decades

The VC industry didn’t always have a consensus on “backing people.” Tom Perkins at Kleiner Perkins looked at technology — is it proprietary? Is it significantly better than the alternatives? Don Valentine at Sequoia looked at markets — a big enough market, he believed, could carry even a mediocre team. His mid-1980s investment in Cisco happened precisely because peers thought the team was weak, but he was convinced the networking market would lift any decent team. Arthur Rock looked at people — he believed a great management team could navigate any specific opportunity.

Rock’s version won the culture war. The reason isn’t complicated: founder-centric narratives make for better marketing. For funds that are essentially selling capital to founders, “we believe in you“ is exactly what founders want to hear.

But “backing people” in today’s practice is no longer what Rock meant. A 2016 survey by Gompers, Gornall, Kaplan, and Strebulaev — covering 681 VC firms and 885 investors — found that 96% listed the team as a key factor, 56% considered it the single most important factor, and 17% of early-stage investors admitted they don’t use any financial metrics at all. An industry overwhelmingly reliant on qualitative judgment, yet without any clear articulation of what the criteria are or how to trace them back to outcomes. Scholars Andrew Zacharakis and G. Dale Meyer put it more bluntly: investors don’t actually understand their own decision-making process.

The Historical Tab for “Backing People”

Diag Davenport, an economist at Chicago Booth, ran the numbers on this “fuzzy humanism” in a 2022 study. Using a dataset of over 16,000 startups and more than $9 billion in committed capital, he trained machine learning models using only information available to investors at the time of their decisions, and asked one question: of the deals VCs actually funded, how many could have been identified ex ante as worse than simply putting the same money into public markets?

The answer: roughly half.

If you removed that half and redirected the capital to a public market index, the overall sample returns would have improved by 7 to 41 percentage points — representing over $900 million in avoidable losses.

The more telling finding was Davenport’s parallel experiment. He trained two models side by side: one to predict the best investments, one to predict the worst. The two models relied on entirely different signals. The model predicting great investments leaned heavily on product characteristics. The model predicting terrible investments leaned heavily on founder backgrounds. In other words: when investors made good decisions, they were actually studying the product. When they made bad decisions, they were studying the team even harder.

This isn’t news to the VC industry. But few people are willing to say it out loud: the more energy VCs spend on “backing people,” the more systematically — and expensively — they get it wrong.

Pricing Foundation Models by Résumé

Map this dynamic onto the global AI investing boom since 2023, and it reads like a controlled experiment. China and the U.S. ran nearly identical playbooks, almost in sync.

On the China side, there was Lightyear (光年之外). In early 2023, Wang Huiwen — known as “Meituan’s number two” — announced he was entering the foundation model space. No product. No complete team. He didn’t write code himself. Yet within months, he raised roughly $230 million at a $1 billion valuation, backed by top-tier USD funds and leading PE firms. The money wasn’t invested in a model or any concrete product thesis. It was invested in the three characters of his name. Months later, he stepped down for health reasons, and the company was absorbed back into Meituan. End of story.

On the Silicon Valley side, there was Inflection AI. Mustafa Suleyman (DeepMind co-founder) + Karén Simonyan + Reid Hoffman — that lineup secured $1.3 billion in funding at a $4 billion valuation, the largest early-stage AI round in North America that year. A year later, Microsoft executed a “non-acquisition acquisition,” pulling out the core team almost wholesale, leaving behind a corporate shell that was repurposed into a B2B API business. Adept and Character AI followed nearly identical scripts through 2024, with Amazon and Google as the respective acquirers.

Line these names up and the conclusion is hard to avoid: among the largest, most hotly pursued early-stage AI checks of 2023, a significant portion were not investments in companies at all. They were risk exposure management on a set of résumés. Even before the outcomes materialized, these deals would have been flagged red in Davenport’s model — their entire thesis was a CV.

The more systematic version played out in each market’s “foundation model” cohort. In China, it was the “AI Six Tigers” (六小虎): Moonshot AI, Zhipu, Baichuan, MiniMax, StepFun, and 01.AI. In the West, it was the ring beyond Anthropic and xAI: Mistral, Cohere, Reka, AI21, and others. What these companies had in common wasn’t product form, technical approach, or market positioning — it was the credential layer of their founders: Google Brain, DeepMind, OpenAI, Meta AI, Tsinghua, CMU, Microsoft Research. Within roughly a year, the vast majority crossed the $1 billion valuation threshold — most while product form was far from stable and business models remained unproven. The investment logic was remarkably candid: the people who “look like they should be able to build foundation models” are these people. Résumés are a reasonable weak signal. But they were being priced as a strong one.

The Dark Horses Outside the “Expected” List

Outside this résumé-driven curve, the most globally recognizable AI companies of the past two years happen to be the ones that least fit the “résumé-pricing model.”

In China, it’s DeepSeek. Founder Liang Wenfeng came from the quantitative hedge fund High-Flyer — not a typical Big Tech AI lab pedigree, and not on most investors’ initial mental list of “foundation model founders.” Yet DeepSeek, through V2, V3, and R1, pushed open-source model capabilities to the global frontier, even forcing Silicon Valley to recalibrate its assumptions about model costs, capabilities, and valuation curves.

In Silicon Valley, it’s Midjourney. Founder David Holz came from the world of haptic interaction, not mainstream AI research. The company took no VC money, did no fundraising PR, yet built one of the most successful consumer-facing AI businesses of this entire cycle on the strength of product and distribution alone.

And then there’s Cursor — not a “model lab alumni” narrative, not a founder-résumé media story, but a team that cut into a specific developer workflow and built an AI coding tool with genuinely compelling product experience.

These companies aren’t counterexamples to “backing people.” Quite the opposite — they demonstrate that the people truly worth backing are often not the ones with the shiniest résumés, but the ones who reveal distinctive judgment within a specific problem.

If the “back people” logic holds, what investors should actually be evaluating isn’t where someone came from, but:

• Why did they choose this problem?

• How do they define this problem?

• What counterintuitive but correct trade-offs have they made in the product?

• Did they see an opportunity the market hasn’t correctly named yet — and see it earlier than everyone else?

Strip away these questions, and “this person is strong” isn’t backing people. It’s backing labels.

What “Backing People” Actually Looks Like

Peter Thiel’s investment in Mark Zuckerberg is often simplified into “early talent recognition,” but contemporaneous accounts show that what Thiel actually valued wasn’t Zuckerberg as an abstract “person” — it was how he chose to approach the problem of online identity, and the specific form Facebook had already taken. Thiel himself said it plainly at a DealBook conference: he doesn’t evaluate people, ideas, business strategy, and technology separately. It’s a complex whole.

Nabeel Hyatt at Spark Capital makes this even more operational: he never decides whether to write a check by watching a founder present. Instead, he uses the product repeatedly, studies every decision embedded in it, and reverse-engineers the judgment, priorities, and problem-solving approach of the person behind it. Sam Altman, during his YC era, formalized it further as the “non-obvious brilliance of an idea” — not whether the founder is smart, but whether the idea is smartly choosing an underestimated problem. Mattia Bianchi at the Stockholm School of Economics and Roberto Verganti at Politecnico di Milano arrived at the same conclusion from the academic side: the most important creative act in entrepreneurship is not solving problems, but defining them.

Put these voices together and the conclusion is clear: the atomic unit of an investment is neither the founder nor the idea — it’s the fusion of the two. An investor who can’t discuss the team without discussing the problem is either investing in shallow pattern-matching or personal charisma. Both will reliably produce Davenport’s bottom half of “predictably bad investments.”

What’s Actually Scarce in AI Investing

In this AI cycle, capital is the last thing the market lacks. Nor is there a shortage of star founders being chased by that capital. What’s genuinely scarce is a more fundamental capability: the judgment to evaluate founders and problems as an integrated whole.

That’s the core reason SVTR continues to build its AI venture ecosystem.

What SVTR aims to document isn’t a collection of AI founder résumés, but a series of founder × problem × product form combinations.

Our AI venture database doesn’t just track founder backgrounds — it tracks the problems companies are tackling, product evolution, funding progress, competitive positioning, and capital relationships.

Our AI venture commentary doesn’t just chase headlines. It tries to answer: Why does this problem matter right now? Why might this company become a new variable? Why is the capital market repricing this direction?

And our AI venture events and programs aren’t about putting founders and investors in the same room. They’re about making sure that people with genuine problem judgment, product judgment, and industry insight get seen earlier and connected more precisely.

Going forward, we want it to be more than a database. We want it to be a signal source for AI investing:

• Helping founders get discovered by the right capital and industry partners.

• Helping investors identify teams with real problem judgment earlier.

• Helping industry players systematically understand AI startups’ technology, product, and commercialization positions.

We believe the next truly important AI companies may not emerge from the most glittering résumé lists. They’re more likely to come from people who are already seriously defining problems, seriously building products, and seriously navigating real markets. What we want to do is make those signals visible — sooner.

Description

​Chinese AI companies are entering a new phase of global expansion.

​The challenge is no longer only about building great products. It is also about building global trust, brand positioning, distribution channels, partnerships, and long-term market presence.

​SVTR and Zhiwai are hosting an intimate private dinner in Palo Alto for a small group of AI founders, investors, GTM leaders, creator-economy builders, and industry partners.

​This closed-door conversation will bring together leaders working across AI, technology, capital, media, and global go-to-market to discuss how Chinese and Asia-born AI companies can better reach global markets.

​The dinner is designed for candid exchange, high-signal networking, and meaningful conversations around AI globalization, brand building, and cross-border growth.

​What We’ll Discuss

• ​How Chinese AI companies can build global brand trust

• ​AI go-to-market strategies in Silicon Valley and beyond

• ​Creator economy as a global distribution channel

• ​Cross-border partnerships between founders, investors, and industry partners

• ​How Asia-born AI products can better position themselves in global markets

​Event Details

​Date: Monday, May 18, 2026
Time: 12:00–3:00 PM
Location: 151 California Ave, Palo Alto
Format: Private dinner / closed-door conversation
Capacity: Limited to 10 seats
Access: Invitation only. Registration is subject to approval.

​Who Should Attend

​This dinner is designed for AI founders, GTM leaders, investors, creator-economy entrepreneurs, and strategic partners who are actively thinking about global expansion, cross-border distribution, AI adoption, or ecosystem collaboration.

​Note

​Seats are intentionally limited to keep the conversation focused and meaningful. All registrations will be reviewed before approval.

The underlying logic is this: as model capabilities commoditize rapidly, the defensibility of the AI application layer is migrating from “intelligence” to “operational depth.” A clean, well-defined, easily standardized workflow means any player with access to the same model can replicate what you do within six months. A workflow stuffed with exceptions, compliance requirements, manual approvals, and legacy system integrations demands that a company spend years rebuilding its organization, data pipelines, and processes from scratch. That rebuilding, in itself, is the new moat.

I. Clean Workflows Have No Future

Over the past two years, Vertical AI has been one of the highest-consensus themes in Silicon Valley venture circles. But companies flying the same “Vertical AI” banner are splitting into two paths with entirely different destinies.

The first type sells a vertical AI product: injecting intelligence into a specific task — faster contract review, more accurate medical summaries, smarter customer service replies. These products demo beautifully, deploy quickly, and can reach $5–10M ARR faster than any prior generation of SaaS. But their problem is structural: when a task is clean enough, the risk low enough, and the integration with existing systems loose enough, it means a model upgrade lets competitors replicate you; a slightly technical customer builds it in-house; and if a frontier lab decides to enter, they simply overwrite you. The purer the task, the faster the replacement.

The second type is doing something fundamentally different. They are not selling “task intelligence” — they are taking over the entire workflow end-to-end: from data ingestion, system integration, approval processes, exception handling, and compliance frameworks, all the way to capturing the full budget a customer used to spend on human labor. On the surface, these companies appear to be building “AI for X.” In substance, they are rebuilding the operational chassis of an entire vertical industry. The former will be compressed into a feature within two to three years. The latter will become the infrastructure of their industry within ten.

What determines the fate of a Vertical AI company is not how smart its model is, but how deep into the dirty work it is willing to reach.

II. The Dirty Work Is the Moat

Using AI to handle auto loan collections doesn’t sound like a sexy thesis. But Salient is a company worth examining closely. What it does is deploy AI voice agents to call delinquent auto loan borrowers — a scenario locked down by three overlapping regulatory frameworks: the FDCPA, TCPA, and Reg F. A single violation can trigger a regulatory investigation. The AI must process federal and state regulations that overlap in real time across jurisdictions, negotiate payment arrangements during the call, comply with contact frequency limits, and know exactly when to hand the call back to a human agent. This is not “a task.” It is a tangled mess of operational, compliance, and technical problems all fused together.

But it is precisely this mess that lets Salient run a business where a human call costs over ten dollars and an AI call costs a fraction of that — and no one can easily take it away. Frontier labs won’t crouch down to chew through compliance and exception handling. Customer companies lack the engineering capacity to build production-grade AI in-house. Competitors, even with access to the same model, cannot bypass the thousands of edge-case workflows Salient has already accumulated.

The same logic keeps surfacing in corners that traditional VCs disdain:

• Medical pre-bill auditing: Charta Health navigates the labyrinthine relationships between payer rules, CPT codes, and denial patterns.

• Freight logistics dispatching: HappyRobot, Pallet, and Augment are replacing the endless phone calls and email coordination among carriers, shippers, and warehouses.

• Tax research: Blue J has entered a $145 billion market populated by 46,000 accounting firms, the vast majority with fewer than ten employees.

These markets are overlooked for two reasons. First, the workflows are dirty. Dirty means resistant to abstraction, resistant to standardization, resistant to the top-down imposition of “general AI capabilities.” Second, the market structure is fragmented. Buyers are scattered across thousands of small-to-midsize operators who generally lack internal engineering teams, lack architectural judgment, and lack the instinct that “technology should be built, not bought.”

These two characteristics are two sides of the same coin: the very thing that makes a market overlooked is also what makes it defensible.

III. TAM Isn’t Underestimated — It’s Measured with the Wrong Ruler

The deeper issue is that these markets aren’t actually small. They’ve just been measured with the wrong yardstick.

The mainstream VC approach to TAM estimation looks at how much an industry spends on software. Apply that ruler to “AI for property leasing,” “AI for medical billing,” or “AI for trucking dispatch,” and the numbers don’t look venture-scale. But this is a fundamentally misaligned ruler — in these markets, the money isn’t in the software budget. It’s in the labor budget, the outsourced services budget, the operational costs distributed across countless small positions.

A property management company might spend $30,000 a year on leasing software, but $300,000 on leasing staff salaries. When an AI product evolves from “assisting leasing agents” to “replacing leasing agents,” its sales target shifts from that $30,000 budget line to the $300,000 one. Push further — when the product expands to maintenance coordination, rent collection, and AI-guided tours — it reaches the entire property operations budget: a number north of $1 million. Same customer, same company; merely by deepening the product, the addressable budget expands by a factor of thirty.

EliseAI is the textbook case. It started as a leasing automation product with a $50,000 ACV — looking every bit like a ceiling-locked proptech play. But as it evolved from “assisting leasing” to “replacing leasing,” then extended into maintenance, rent collection, and AI-powered tours, it now serves one-eighth of all apartment units in America, with per-customer annual spend reaching the million-dollar range. The company itself has begun replicating the same playbook into healthcare’s $600 billion administrative cost pool. The market didn’t get bigger. The product revealed how big the market always was.

This mechanism also explains a critical paradox: why these markets can accommodate $10-billion companies yet won’t be directly targeted by OpenAI or Anthropic. Viewed through the software-spend lens, the market is too small for a frontier lab to justify mobilizing massive resources. Viewed through the labor-spend lens, the market is already large enough — but by the time it surfaces in public data as an “obvious opportunity,” the vertical player who has been grinding for five years will have already stacked up the data, integrations, and distribution.

The profitability model and organizational structure of frontier labs are simply not built for this work. They must continuously invest in the model frontier, must maximize token revenue (which directly conflicts with customers’ interest in moving toward agentic workflows), and cannot possibly build custom-grade applications for dozens of verticals in parallel. In most vertical markets, the AI application layer won’t be consumed top-down by labs. It will be won by vertical players willing to dig deep and trade focus for advantage.

IV. Implications for AI Investing

This thesis carries unmistakable implications for founders and investors navigating the AI wave.

For founders, the most dangerous temptation is to build something that “looks like AI”: a polished demo, a clean SaaS interface, a tidy workflow automation. These products will reach $5M ARR the fastest and get compressed into a feature the fastest. What’s truly worth building lies in directions that seem, at first glance, absurdly labor-intensive, hopelessly fragmented, and populated by buyers so unsophisticated you wonder if they’ll even sign a contract. It demands a temperament entirely different from the past decade of SaaS entrepreneurship: a willingness to make part of the company an “operations company” rather than purely a “software company,” a willingness to shoulder the dirty work of compliance, exceptions, and service delivery, and a willingness to accept lower margins than pure software in the early days in exchange for deeper customer embeddedness.

For investors, the screening mechanism needs an upgrade. The center of gravity in AI investing is shifting from “can the model do this?” to “is the workflow ugly enough?” A Vertical AI company worth backing should not have its TAM estimated by “software spending in the market” but by “the total labor and service costs the industry spends on this function.” A founder worth backing should not merely demonstrate the boundaries of model capability but the depth of their understanding of the industry’s dirty work: which exceptions account for 80% of the workload, which compliance requirements have been the Waterloo of every software company for the past decade, and which integrations truly prevent competitors from replicating.

It’s worth noting that this playbook carries an underappreciated divergence between the U.S. and China. The reason “replacing labor” drives such dramatic TAM expansion in the American market is that the multiplier between software budgets and labor budgets is enormous — per-customer labor spend is often ten times or more the software spend. China’s domestic labor cost structure compresses that multiplier, making per-customer expansion potential naturally narrower. But in reverse, Chinese teams have a natural lack of resistance to “doing operations, doing delivery, doing service” — precisely the part Silicon Valley founders are least willing to touch. We’re already seeing, within SVTR’s network, a cohort of Chinese-background teams bringing this operational resilience into America’s fragmented vertical markets. They may not have the strongest models, but they’re willing to go two steps further on the dirty work than their Silicon Valley competitors. In the Vertical AI generation, that willingness itself is the scarcest asset.

This is also the direction we continue to track: which industries have the most complete triangle of “heavy labor + high fragmentation + low tech DNA,” which teams truly understand the difference between “selling software” and “selling a system,” and which opportunities offer structural arbitrage between the U.S. and Chinese markets. The next batch of breakout companies will look, on the surface, less “AI” than many Silicon Valley darlings — but they will quietly take over the operational chassis of an entire industry.

Models win the demo. Wedges win the pilot. Systems win the market. The real bet in Vertical AI was never on the smartest end — it’s always been on the end most willing to do the dirty work.

This is not a sampling error. SVTR’s AI Deal Database, tracking the past eight weeks of data, shows that among rounds of $100 million or more, the share held by general-purpose AI application companies has fallen from 38% in Q3 2025 to 0% last week. Over the same period, the median valuation for research-stage seed rounds rose 2.4×, and for late-stage vertical Agent rounds, 1.7×.

The collapse of the middle layer deserves more attention than the boom at either end. (See chart at the end: Last Week’s Top 10 Funding Rounds.)

Signal 1 | Seed Rounds Are Being “Unicornified” by Paradigm Research

David Silver’s Ineffable Intelligence redefined what “early-stage betting” means with a single $1.1 billion seed round at a $5.1 billion valuation. Sequoia, Lightspeed, Index, Google, and Nvidia all came in simultaneously — a cap table that looks less like a typical seed syndicate and more like a Series C investor lineup transplanted to day one.

This wasn’t an isolated case. The same week, Standard Intelligence closed a $75 million seed at a $500 million post-money valuation, betting on computer-use models. On the China side, Wujie Power accumulated over $200 million in angel funding on a “native world model + reinforcement learning” thesis. What these three companies share isn’t a sector — it’s a narrative: they’re all betting on “post-LLM paradigms.”

When capital is willing to assign a $5 billion valuation on day one, it means the LP base has already accepted a new premise: the capability curve of general-purpose large language models is converging. The next leap won’t emerge from fine-tuning or RAG — it will require entirely new training paths (autonomous RL, world models, computer use). Once this premise takes hold, the scarcity of “founder credentials” gets rapidly financialized, while the moat narrative for downstream general-purpose application layers gets diluted in reverse.

Signal 2 | Vertical Agents Have Stopped Telling Growth Stories — They’re Telling Category Lock-In Stories

The back half of last week’s Top 10 features a dense cluster of large late-stage rounds forming the other pole: Rogo (investment banking AI / $160M Series D / $2B valuation), Aidoc (medical imaging / $150M Series E / 31 FDA clearances / 1,600+ hospitals), Hightouch (marketing data / $150M / $100M+ ARR / $2.75B valuation), Avoca (home services AI / $125M+ / 800+ enterprise clients), and Netomi (customer service / $110M / clients include Delta and the NBA). Add Legora (legal AI / $50M / $100M+ ARR / $5.6B valuation) from the same week, and that’s at least six companies exhibiting “category leader” characteristics within their respective vertical workflows.

These companies differ from “vertical AI startups” of twelve months ago in two critical dimensions. First, ARR is no longer a $10 million-level story — it’s a $100 million-level reality. Second, the valuation anchor no longer derives from “market size × penetration rate” but from “workflow embedding depth × customer switching cost.” Hightouch’s $2.75 billion valuation against $100M+ ARR implies a 27× multiple — a figure that, in SaaS history, corresponds to companies with established category leadership at Series E/F, not growth-stage companies that just crossed the $100M ARR line.

When capital is willing to pay that multiple, it means these companies are already being priced as the next generation of ServiceNow or Salesforce within “vertical SaaS + AI” — not as one more variant among the sea of general-purpose Copilots.

Signal 3 | The Middle Collapses: General-Purpose AI Products Are Collectively Absent from the Top 10

Zoom out, and not a single company in last week’s Top 10 is a horizontal Copilot, general-purpose Agent platform, or general-purpose LLM application-layer play. The closest candidate to a “middle layer” company might appear to be Parallel Web Systems ($100M Series B / $2B valuation), founded by former Twitter CEO Parag Agrawal — but the product is explicitly positioned as “web search infrastructure for AI agents,” serving downstream Agents as infrastructure rather than being an Agent product itself.

This is a quiet turning point. From 2024 through the first half of 2025, general-purpose Copilots, general-purpose Agent orchestration platforms, and horizontal LLM applications were the most crowded lanes in the AI primary market. Multiple companies secured $1 billion+ valuations at sub-$10M ARR. Last week, the absence of such companies from the Top 10 isn’t a valuation markdown — it’s a complete exit from the frame.

Why is the middle collapsing? Because it’s being squeezed from both directions simultaneously. From above, OpenAI, Anthropic, and Google are embedding general-purpose Agent capabilities directly into their flagship products (ChatGPT, Claude, Gemini), making it nearly impossible for Copilot startups to build moats beyond the model itself. From below, vertical Agent companies have locked in paying enterprise customers through workflow depth, keeping the switching cost of replacing a “general-purpose product” on the procurement list extremely low. When both directions are squeezing, capital simply stops betting on the middle.

Root Cause | Why Now

This repricing is, at its core, a second-order reaction to capability convergence at the LLM model layer.

In 2024, “model capability” itself was still a differentiating variable. General-purpose products could establish temporary leads through prompt engineering, RAG, or fine-tuning. By the second half of 2025, the gap between flagship models on general tasks was narrowing rapidly (GPT-5, Claude Opus 4.7, and Gemini 2.5 shuffle their rankings on standard benchmarks week to week), meaning “model capability delta” can no longer sustain a $1 billion valuation as a moat.

Under this premise, capital can only find genuine differentiation at the two extremes: either bet on the next paradigm (one that doesn’t sit on the LLM convergence curve), or bet on vertical workflows (which don’t compete directly on model capability). The middle layer is on neither the new paradigm nor the workflow-depth axis — so its pricing loses its anchor.

This also explains the tight temporal clustering of all three signals last week: they aren’t independent events, but the same underlying logic surfacing simultaneously across different cross-sections.

Implications | Who Gets Caught in This Repricing

For founders: The fundraising window for general-purpose AI application startups is closing fast. If you’re building a horizontal Copilot, a general-purpose Agent platform, or a prompt-engineering-based LLM application-layer product, you’ll need to make a choice within the next 12 months. Either converge your product depth toward a specific vertical workflow (healthcare, legal, finance, customer service, marketing — pick one), and let ARR, not user counts, do the talking. Or pivot toward a position closer to the model/data/training layer. Products stuck in the middle will struggle to raise their next round on the back of a “fast growth” narrative alone. Over the past 60 days, SVTR’s AI Deal Database shows that 100% of application-layer companies raising rounds of $50 million or more were vertically positioned.

For investors: Valuation frameworks need a rewrite. The old logic of valuing general-purpose AI applications through “DAU × ARPU” or “user growth rates” is breaking down, because these metrics no longer correspond to sustainable moats at the general-purpose layer. The new anchors need to return to two variables: on the vertical axis, “workflow embedding depth” (switching cost, regulatory barriers, data flywheels); on the model axis, “training-paradigm scarcity” (research credentials in RL, world models, computer use, and other non-LLM mainline directions). If a portfolio company has neither accumulated workflow depth nor placed a bet on a new paradigm, its valuation anchor needs to be proactively adjusted downward.

For Big Tech: OpenAI, Anthropic, and Google’s “product-layer squeeze” strategy was indirectly validated by last week’s capital flows. Their incentive to keep squeezing upward at the general-purpose layer will only strengthen.

For enterprise buyers: On procurement lists, the budget split between “general-purpose Copilots” and “vertical workflow Agents” will continue to diverge. The former increasingly folds into Big Tech subscriptions; the latter becomes a standalone line item.

Validation | Three Key Metrics Worth Tracking

Over the next 90 days, these three metrics can validate whether this signal holds:

1. Median seed valuation for general-purpose AI application-layer companies. Watch whether the median seed valuation for horizontal Copilots and general-purpose Agent platforms in Q1 2026 has declined 30%+ from Q3 2025. If so, the repricing thesis is confirmed.

2. Number of vertical AI companies entering the $2B+ valuation tier. Watch whether, within 90 days, three or more vertical AI companies with ARR above $50M enter the $2B+ valuation zone. If so, the category lock-in logic is being consistently priced in.

3. Whether a second “seed-round unicorn” appears. Watch whether another company with RL, world model, or computer-use research credentials emerges at a $500M+ seed valuation. If two or more appear, “academic-founder unicornification” has become a structural phenomenon rather than an outlier.

If all three metrics are confirmed within 90 days, last week’s Top 10 reflects not a weekly fluctuation, but the starting point of the 2026 AI primary market pricing framework.

SVTR’s AI Deal Database has completed retroactive tagging across three categories — “model research layer / general-purpose product layer / vertical workflow layer.” It can now filter comparable historical rounds and valuation multiples across the six verticals featured in last week’s Top 10 (investment banking, medical imaging, marketing, home services, customer service, legal) over the past 72 months, providing a reference map for founders and investors recalibrating their pricing frameworks this quarter.

That’s how many of last week’s largest AI funding rounds went to companies building things you can touch — robots, chips, reactors, trucks, energy systems. The top 10 companies tracked by the SVTR AI Venture Database raised nearly $10 billion in combined fresh funding (excluding Project Prometheus’s $10 billion round still in negotiation). The lone exception, VAST Data, counts labs training large physical models among its key clients.

Meanwhile, the hottest pure software Agent startups? They raised between $5 million and $20 million per round. Two orders of magnitude smaller.

This isn’t an AI downturn. It’s a reallocation of capital weight. And if you’re a founder, investor, or enterprise leader in the AI space, the implications are enormous.

What the Money Is Actually Saying

Let’s start with a simple exercise. Add up all the representative software Agent rounds from last week: Orkes ($60M), NeoCognition ($40M), Petual ($20M), Band ($17M), Coral ($12.5M), Brev ($3.3M), Zig.ai ($3M). The total comes to less than $150 million.

That’s roughly equal to the single round raised by Reliable Robotics — a company that flies cargo planes.

Let that sink in.

The capital isn’t just flowing toward physical AI. It’s flooding in across every layer of the stack, forming what looks like a complete industrial architecture for the first time.

The Three-Layer Stack of Physical AI

Think of physical AI as a three-layer system, and last week’s deals lit up all three independently:

The Simulation & Design Layer turns physical processes into computable, optimizable objects. Project Prometheus (in talks for $10B), CuspAI ($200M), and others are building the digital twins and design engines that make physical AI possible. These companies are often misclassified as “software” — but their customers pay based on manufacturing line and lab outcomes, not SaaS subscriptions.

The Compute & Infrastructure Layer provides the chips, data pipelines, and networks optimized specifically for physical AI workloads — not general-purpose GPU training. VAST Data ($1B Series F), EVAS Intelligence (CNY 1.5B), Sunrise (AI inference GPUs), NineTree (neuromorphic chips), and others are building this middle layer.

The Body & Application Layer is the most visible destination for capital. Last week alone, at least five humanoid and industrial robotics companies secured funding: Blue Energy ($380M), DeepWay ($310M Pre-IPO), Reliable Robotics ($160M), Pudu Robotics ($150M at ~$1B valuation), and Booster Robotics (nearly $100M).

Here’s what matters: these three layers now form a closed industrial chain. The simulation layer tells the compute layer “how to build this reactor.” The compute layer tells the body layer “how to drive this truck.” The body layer feeds real-world data back to the simulation layer.

A year ago, this chain was broken. Last week, it connected for the first time in terms of funding rhythm.

Signal 1: Physical AI Deals Are 10x Larger Than Software Agent Rounds

When you split last week’s funding by whether a company delivers physical actions or products, the gap is impossible to ignore:

• Physical AI: Median deal size ~$160M per round. Strip out mega-rounds (Prometheus, VAST), and the median still holds at $150M.

• Software Agents: Median deal size ~$17M per round. Even the largest round (Orkes, $60M) is smaller than the smallest physical AI round in the top 10.

The gap isn’t 2x or 3x. It’s more than an order of magnitude.

Capital no longer believes the “small check, massive upside” logic holds broadly for Agent startups. The market is repricing in real time.

Signal 2: A Complete Stack, With Independent Funding Across All Layers

If concentrated funding were limited to isolated outliers, it wouldn’t constitute a trend. But last week’s data shows full industrial depth. Capital is flowing independently into simulation, compute, and body layers — each attracting its own dedicated investors and its own deal logic.

That’s what separates a trend from an anomaly. When multiple layers of an industrial chain attract capital simultaneously and independently, you’re looking at a structural shift, not a hot sector.

Signal 3: The “Zero Marginal Cost” Narrative for Software AI Is Hitting a Wall

The core investment thesis behind software AI was elegant in its simplicity: model capabilities would keep breaking through, marginal costs would approach zero, and winners would take all. This narrative supported the sky-high valuations of OpenAI, Anthropic, and Mistral, as well as the “grab the entry point first, build the business later” logic of Agent startups.

That narrative began hitting three boundaries simultaneously in the second half of 2025:

First, the pace of generational leaps in model capability has slowed. Each new model delivers noticeably smaller increments over its predecessor.

Second, Agent startups have become overly homogenized. “GPT-wrapper plus workflow” products went from scarce to template within six months, blurring valuation anchors.

Third, inference costs are not approaching zero. GPU compute has become a hard constraint, making the marginal cost of one API call increasingly close to the marginal cost of delivering one physical service.

Capital won’t wait for the narrative to collapse naturally — it migrates first. And migration destinations must meet two conditions: unit economics must be measurable, and external barriers must not be controlled by large model companies.

The physical world meets both criteria perfectly:

• The unit gross margin of a robot, a truck, or a reactor is crystal clear.

• The data generated from motion, force feedback, and environmental closed loops is currently inaccessible to OpenAI and Anthropic.

• Chinese manufacturing capacity spillover and U.S. reshoring policy tailwinds create a rare window where both cost curves and policy incentives align.

This is the real pricing logic behind Project Prometheus’s $38B post-money valuation. Bezos isn’t investing in the next OpenAI. He’s investing in the AWS of the physical world — a base layer providing simulation and design infrastructure for all downstream vendors building physical AI.

Why Now?

The timing is no accident. Three forces are converging in the same window:

1. Scaling diminishing returns — Large models have hit marginal diminishing returns simultaneously across algorithms, compute, and data for the first time.

2. Hardware cost decline — U.S. manufacturing reshoring policies are overlapping with Chinese robotics production capacity spillover, pushing hardware BOM costs into a rapid decline channel.

3. GPU capex repricing — The market’s patience for “compute-consumption” business models has shortened dramatically.

Who Should Care

If You’re a Founder

Here’s the one question that matters: Does your product have a closed loop from physical action to physical result?

If not — if what you’re building can be summarized as “adding an Agent workflow on top of a large model” — last week’s data suggests your valuation ceiling is likely to keep falling in 2026, not rising.

Conversely, if you can embed AI into a specific physical scenario (a robot joint, a factory production line, a design process step), even a small team is more likely to secure patient capital.

One telling data point: companies that raised more than $100M last week have an average founding year of 2020. The market is willing to pay a premium for teams that have put in five years of hard work.

If You’re an Investor

Time to shift your investment paradigm from SaaS templates back to semiconductor and new energy templates. Physical AI rounds are characterized by large check sizes, long cycles, strong industrial chain ties, and less predictable exit timelines.

Funds that previously specialized in ARR models need to build three new due diligence capabilities:

• Hardware supply chains and manufacturing processes

• Energy and material structural bottlenecks

• Cross-border capacity scheduling

Watch out for a valuation trap: early-stage physical AI valuations are already rising on “strong storytelling,” but it will take 18-24 months to see if unit economics actually work. Don’t apply software AI pacing to these deals.

If You’re Big Tech

• Bezos is building a new flagship outside the OpenAI ecosystem via Prometheus.

• NVIDIA is locking in the compute consumption side of physical AI through investments in VAST and others.

• Microsoft and Google’s relative positions are becoming more nuanced — and potentially more vulnerable.

If You’re an Enterprise Customer

Manufacturing, logistics, energy, and medical devices are the four industries most likely to capture the first wave of “physical AI application dividends” in 2026. If you’re in one of these sectors, the time to start evaluating physical AI vendors is now.

Three Numbers to Watch Over the Next 90 Days

To confirm whether last week’s trend is structural or noise, track these three indicators:

1. The physical-to-software ratio at the $100M+ level. Count how many physical AI companies raise >$100M rounds vs. pure software Agent companies crossing the same threshold. If the gap widens to 5:1 or higher, this is not a one-off anomaly.

2. NVIDIA and TSMC partnership announcements. If 3+ partnerships explicitly target “physical simulation, robotics, industrial AI,” the upstream of the industrial chain has formally tilted.

3. VC thesis language. Track the frequency of “Physical AI,” “Embodied AI,” and “Physical Simulation” in theses or annual letters from top U.S. and Chinese VCs. If it doubles compared to the first half of 2025, LP consensus has formed and capital will concentrate further.

Last Week’s AI Funding Leaderboard

The SVTR AI Venture Database has updated complete data cards for all 6,000+ tracked companies, adding a new “Physical AI” tag. Users can filter more than 30 active targets by the three-layer structure for cross-sector comparison.

Let’s get one thing out of the way: the hardest unsolved problem in the agent economy is not making agents smarter. Models will keep improving — that’s the easy bet. The hard problem is everything around the model: the payment rails, the security layers, the trust protocols, the interoperability standards. Intelligence without infrastructure is a demo. Infrastructure is what turns it into a business.

Everyone’s talking about how smart AI agents are getting. But the real story right now isn’t about what agents can do — it’s about what they can’t do yet because the infrastructure doesn’t exist.

Payments. Security. Identity. Interoperability. These aren’t glamorous problems. They’re plumbing. And plumbing is exactly what separates a demo from a production deployment.

Over the past 12 months, we’ve tracked a new category of companies emerging in our AI venture database: agent middleware. A year ago, this category barely existed. Today, there are over a dozen funded startups building the connective tissue that agents need to operate in the real world — and the investor rosters now include Kleiner Perkins, a16z Scouts, Polychain.

This isn’t a coincidence. This is a sector being systematically priced in.

While embodied AI and coding agents continue absorbing large rounds, the consensus there is already high. The real “low consensus, high signal” space? Agent middleware.

Signal 1: Payment Rails — Agents Need Their Own Financial Infrastructure

On the surface, you might think existing payment APIs — Stripe, bank integrations — should work fine for agents. They don’t. Here’s why:

When the entity initiating a transaction shifts from a human to a machine, the entire trust model breaks. Traditional fraud detection is built around stolen credentials and unauthorized card swipes. But an agent operates under delegated authority — it can execute multiple transactions across environments without any human interruption. The attack surface moves from the checkout page to the system architecture itself.

The market has noticed. In the past 90 days alone:

• Ralio (London, $2.5M seed) — payment infrastructure for agents with built-in risk controls and audit trails

• Nava (New York, $8.3M seed) — verification, guardrails, and custodial execution for agent-managed capital

• Natural ($9.8M seed, late 2025) — explicitly positioned around “agentic payments”

As FinTech Weekly noted in April 2026: the market is already putting a price tag on the question of “can an agent legally spend money?” That’s the first hard prerequisite for the agent economy to move from experiment to production.

And it naturally leads to the next question: once an agent spends money, how do you know it hasn’t been hijacked?

Signal 2: Security & Auditability — The “Insider Threat” Problem, But for Agents

In December 2025, OWASP published its Agentic Applications Top 10 — the first-ever security risk framework specifically for autonomous AI agents, developed by over 100 researchers. The most striking concept: the “rogue agent” — a system that has proper authorization, appears to be functioning normally, but has deviated from its design intent.

This isn’t an external hack. It’s an internal mutation. The agent equivalent of an insider threat.

The industry has moved from “what can agents do?” to “what happens when agents go off the rails?” Capital is responding fast:

• Capsule Security (Tel Aviv, $7M seed) — runtime agent security and trust layers

• ActionAI (New York, $10M seed) — explainable, auditable, rollback-capable execution infrastructure

• Microsoft open-sourced its Agent Governance Toolkit in April 2026

• Ledger announced it’s extending its hardware security architecture from crypto to agent identity verification

And the attack side is already racking up real costs: security researchers disclosed in April 2026 that 26 LLM routers were secretly injecting malicious tool calls and stealing credentials. One case resulted in $500,000 stolen from a client wallet.

If payment rails solve “how does an agent spend money,” the security layer solves “can we trust how the agent spent it.” But even together, they’re not enough — when multiple agents need to collaborate on a task, you need something deeper.

Signal 3: Agent OS & Transaction Protocols — Who Defines the “Handshake” Between Agents?

When multiple agents need to work together, a fundamental question emerges: how do they discover each other, exchange data, and establish trust? No single company’s SDK solves this. What’s needed is a protocol layer — something analogous to HTTP or OAuth, but for agent-to-agent interactions.

Active companies in this space include:

• Keman Intelligence (Hangzhou, ~$10M, building an A2H marketplace; founders from Didi)

• Panfeng Intelligence (Hangzhou, tens of millions RMB, A2A-native e-commerce Agent OS; founders from DingTalk’s open platform)

• Creao AI (Hong Kong, $10M Pre-A, Agent OS platform; backed by Sequoia China and Hillhouse)

• Modern Relay (Barcelona/San Francisco, $3M, shared data layer connecting people, software, and agents)

A pattern worth noting: nearly every founding team comes from major platform companies — Didi, DingTalk, Meta, Amazon. They lived through the last generation of “platform ↔ developer” ecosystem construction, and they’re now trying to replicate that paradigm in an “agent ↔ agent” world.

One signal that’s easy to miss if you only read English-language tech media: Chinese teams are disproportionately active in A2A and A2H transaction protocol layers. This is a structural signal, not a trend piece.

Why Now?

Three timing factors converged in Q1–Q2 2026 to trigger this middleware funding wave:

1. Regulation got real. The EU AI Act’s high-risk obligations take effect in August 2026. Colorado’s AI Act kicks in June 2026. Enterprises deploying agents now must have auditable, traceable execution records — which is exactly what companies like Capsule Security and ActionAI provide.

2. Agent transaction volume became economically meaningful. The World Economic Forum, citing Adobe data, reported that AI-driven retail site traffic surged 805% year-over-year during Black Friday 2025, with agents helping drive over $22 billion in global online sales. When agents start materially participating in money flows, infrastructure needs stop being theoretical.

3. Security incidents went from hypothetical to actual. Poisoned LLM routers, stolen agent wallets — these are no longer red-team exercises. They’re incidents that happened. They’ve forced enterprise buyers and investors to reassess the prerequisites for agent deployment.

Who Should Care

If you’re building agent applications: Ask yourself — what infrastructure does your agent rely on for payments, security, and cross-agent collaboration? Until these pipes mature, your deployment ceiling with enterprise customers will be painfully obvious. Conversely, if you have platform-level experience (especially in payments, identity, or developer ecosystems), the entry window is short.

If you’re investing: Agent middleware is in a “seed-to-Series A” density phase. Valuations haven’t been inflated by application-layer narratives yet. But the window is closing fast — once one or two companies establish network effects in payments or security, late movers will face a much narrower path.

If you’re at a platform company: Microsoft and Ledger are already making moves, but history shows that platforms tend to ship “80% good enough” general solutions. Depth in vertical scenarios will still belong to startups.

If you’re an enterprise buyer: The most important question to ask your agent vendor isn’t “how smart is your agent?” It’s “what does the audit chain look like when your agent initiates a transaction?” and “what’s your rollback mechanism when agent behavior deviates?”

Three Metrics to Watch Over the Next 90 Days

1. Enterprise customer sign-ups for agent payment companies. Watch whether Ralio, Nava, and Natural can each sign 3+ paying enterprise customers within 90 days. This validates whether “agents need independent payment rails” is a real thesis — or whether existing payment infrastructure APIs are good enough.

2. EU AI Act compliance-driven security procurement. Before the August 2026 high-risk deadline, watch for European enterprises bulk-purchasing agent security and audit tools. If Capsule Security or peers show an ARR jump in Q2–Q3, it confirms regulation as the core catalyst for this sector.

3. Series A density in China’s A2A / Agent OS space. Keman Intelligence and Panfeng Intelligence are still at angel stage. If two or more Series A rounds close in this space within 90 days, it signals that China is independently forming its own agent transaction protocol ecosystem — not simply following Silicon Valley’s narrative.

These locations are not simply different urban districts or industrial zones. Each represents a distinct path of AI development within China’s broader innovation landscape. Taken together, they offer a more grounded way to understand Chinese AI—not only through the lens of individual companies or technical breakthroughs, but through the structure of the ecosystem itself.

After one week on the ground, our conclusion was clear: China’s AI competitiveness is increasingly not defined by a single company, a single model, or a single technological leap. It is being shaped by a more layered and coordinated innovation system.

Based on this field visit, SVTR summarizes six emerging models of Chinese AI innovation.

1.

Productization Innovation: AI Is Entering Real Products Faster

At AWE Shanghai, one trend was immediately visible: AI is moving rapidly into consumer products and intelligent devices.

From AI-powered robots and home appliances to vision-enabled hardware and smart terminals, the deployment of AI in China is no longer confined to conceptual demos or software interfaces. It is increasingly embodied in tangible products entering the market at speed.

What this reflects is not just technical progress, but a structural advantage. China’s manufacturing depth, hardware ecosystem, supply-chain coordination, and large domestic market together create a shorter path from prototype to commercialization. In the AI era, that matters enormously.

This means China should not be understood only as an application market for AI. It is also becoming one of the most important environments for AI productization and large-scale market validation.

2.

Park-Driven Innovation: China’s Industrial Parks Function as Systems

Suzhou Industrial Park highlighted a model with distinct Chinese characteristics: park-driven innovation.

In many markets, innovation clusters are formed primarily through spontaneous concentration of talent and capital. In China, industrial parks often play a broader and more active role. They do not simply offer office space. They combine industrial planning, policy support, capital guidance, incubation, and supply-chain coordination into an integrated platform.

For AI startups, the value of this model is practical and immediate. It lowers the threshold between technical capability and commercial deployment. It reduces friction in areas such as customer access, pilot opportunities, regulatory navigation, and ecosystem connection.

In this sense, Chinese industrial parks are better understood not as real estate carriers, but as infrastructure for organizing innovation resources.

One of the strongest impressions shared by overseas founders on this trip was that Chinese AI development is often not just company-led innovation, but regionally coordinated innovation.

3.

Startup Cluster Innovation: Ecosystem Density Still Matters

At Shanghai’s Xuhui ModelSpace, another important dynamic became clear: the efficiency created by high-density startup ecosystems.

In one concentrated environment, we saw AI startups, large-model teams, investors, researchers, and enterprise users interacting at high frequency. This kind of proximity produces real advantages: faster technical exchange, more efficient fundraising conversations, quicker product feedback loops, and stronger alignment between founders, capital, and market demand.

For AI, many of the most important opportunities still emerge through repeated offline interaction rather than purely digital connectivity. Teams that can more quickly access technical talent, customer signals, investor perspectives, and industrial resources tend to move faster in the early stages.

Several overseas entrepreneurs noted that this environment, in some respects, reminded them of Silicon Valley—not because the two ecosystems are identical, but because density itself creates momentum.

AI entrepreneurship is not only a competition of models and products. It is also a competition of ecosystem density.

4.

AI in the Physical World: Robotics and Intelligent Manufacturing Are Becoming Core Directions

At the Shanghai Maqiao Artificial Intelligence Innovation Experimental Zone, another shift came into focus: AI is moving beyond the digital layer and into the physical world.

What stood out there was not only software or model capability, but a broader industrial ecosystem tied to robotics, intelligent manufacturing, logistics automation, and core components. This points to an important next phase of AI development.

If the previous wave of AI was centered largely on information processing and software workflows, the next phase is likely to depend much more on how AI integrates with machines, factories, infrastructure, and real-world operations.

China has clear advantages in this domain. Its manufacturing base, engineering capabilities, supply-chain depth, and large industrial deployment environments create conditions that are difficult to replicate elsewhere.

For that reason, robotics and intelligent manufacturing should not be seen as peripheral application areas. They may become one of the strongest long-term growth directions for Chinese AI in the global market.

5.

Industrial Chain Cluster Innovation: Coordination Across the Value Chain Is a Strategic Strength

Zhangjiang AI Town illustrated another key feature of China’s AI ecosystem: industrial chain clustering.

What makes this model powerful is not dependence on a single standout company, but the co-location of multiple interdependent layers of capability. In one relatively concentrated geography, we observed AI chips, model platforms, robotics companies, smart hardware players, application teams, and supporting services beginning to form a more complete collaborative structure.

The significance of this kind of ecosystem is that it improves both efficiency and resilience. Innovation is no longer driven only by isolated excellence. It is accelerated by upstream and downstream coordination across R&D, testing, manufacturing, supply, channels, and customer integration.

For many international founders, this is one of China’s most compelling advantages. The attraction is not only market size, but the possibility of finding a workable cooperation chain more quickly and within a shorter operating radius.

As AI competition deepens, the real contest may increasingly be not between individual companies, but between ecosystems with different levels of coordination. From that perspective, industrial chain clusters represent one of the most important structural strengths of Chinese AI.

6.

International Innovation Platforms: Lin-gang Is Exploring a Global Interface for Chinese AI

The Lin-gang Special Area represented a more future-oriented model: the international innovation platform.

Compared with the other regions we visited, Lin-gang stood out for its emphasis on openness. Its development logic is closely tied to cross-border capital flows, global talent attraction, international business environments, and policy experimentation around global-facing innovation.

This matters because for overseas founders, entering China has never been a simple market-entry decision. It involves local networks, policy understanding, industrial partnerships, operational support, and long-term trust building. Regions that can absorb and connect international innovators more effectively will play a different role in the next stage of AI development.

In that sense, Lin-gang is not just serving local industrial growth. It is exploring how Chinese AI ecosystems can build more effective interfaces with the rest of the world.

That makes it not only a regional case study, but also a signal of where future institutional innovation may emerge.

7.

A Larger Judgment: China’s AI Advantage Is Becoming Systemic

Viewed together, these six models suggest a broader conclusion.

China’s AI development path is increasingly not about a single “point breakthrough.” It is about systemic innovation capacity.

That system includes the speed of productization, the organizing role of industrial parks, the efficiency created by high-density startup clusters, the integration of AI into physical industries, the coordination enabled by industrial-chain ecosystems, and the global-facing potential of more open innovation platforms.

This has broader implications for the global AI landscape.

A new division of labor may be taking shape. Frontier breakthroughs may continue to emerge from leading research centers around the world. But large-scale productization, industrial deployment, and real-world validation are increasingly tied to ecosystems with manufacturing depth, operational coordination, and policy support.

In that context, China’s role in AI is becoming clearer. Its importance lies not only in market size, and not only in application adoption. It lies in its ability to connect technology, industry, capital, infrastructure, and commercialization into a more integrated system.

The future of AI competition may therefore depend less on who has the strongest standalone model, and more on who can connect the strongest innovation systems.

One of the most important combinations to watch may be this: frontier technical capability from global research centers, combined with China’s industrialization and deployment capacity.

8.

Final Thought

This trip reinforced a view that is becoming harder to ignore: China’s AI competitiveness is increasingly a function of system capability.

Technology, manufacturing, capital, policy, industrial infrastructure, and international connectivity are no longer operating as separate layers. They are increasingly interacting as part of one larger engine.

For global founders, investors, and ecosystem builders, the opportunity ahead may not simply be to identify the next strong company. It may be to better understand how different AI systems work, where their strengths are complementary, and how new forms of cross-border cooperation can be built on top of that understanding.

That is one of the questions SVTR will continue to study through research, field observation, and long-term engagement across global AI ecosystems.

If you are a founder, investor, or ecosystem builder interested in joining SVTR’s global AI venture platform, you can apply here: