Startup Funding Watch: Capital Flows Into the AI Economy’s Core Layer

Startup Funding Watch | September 23, 2026

AI Data Infrastructure and Cybersecurity Dominate as Snorkel AI Triples Valuation to $3.5B and Cyera Raises $1B

Executive Brief

The AI Data Layer Is Now the Funding Market's Center of Gravity

Two massive rounds this week crystallized the clearest funding signal of the month: capital is concentrating around the infrastructure layer that makes AI models trainable, secure, and deployable. Snorkel AI raised $350 million in Series E funding at a $3.5 billion valuation — nearly triple its previous mark — as demand for high-quality training data and reinforcement learning environments explodes across frontier AI labs. Cyera, an Israeli AI and data security company, raised $400 million from Goldman Sachs, extending its June round to a total of $1 billion and maintaining a valuation above $12 billion.

The pattern is not simply that AI is attracting capital. It is that the market is now pricing the invisible layers of the AI stack — training data pipelines, agent security, identity management, and the specialized risk infrastructure that enterprises need before they can deploy autonomous systems at scale. Meanwhile, infrastructure continues to anchor the top of the market: Vantora, the physical-AI startup builder formerly known as UP. Labs, raised $100 million+ from Silversmith Capital Partners.

Funding at a Glance

Company	Amount	Sector	Lead Investor
Cyera	$400M extension	AI Data Security	Goldman Sachs
Snorkel AI	$350M Series E	AI Training Data	Insight Partners, S32
Qupital	$300M Series C	AI Trade Finance	M Capital
Vantora	$100M+	Physical AI	Silversmith Capital Partners
Chamelio	$26M Series A	Legal AI Agents	Entrée Capital

Signal summary: The AI data and security layer now commands more capital than most application categories. Snorkel and Cyera alone represent $750 million in fresh funding — and both are being valued on strategic position, not revenue alone.

Biggest Funding Deals

1. Snorkel AI — $350M Series E at $3.5B

AI Training Data Startup Snorkel AI Triples Valuation to $3.5B as Frontier Labs Chase High-Quality Data

Company: Snorkel AI | Stage: Series E | Amount: $350 million | Valuation: $3.5 billion | Lead Investors: Insight Partners, S32 | Participating: Third Point, March, Blumberg, Allegis, Standard VC, Frontline, Addition, Lightspeed, Greylock, GV, Wells Fargo | What it does: Helps AI labs and enterprises build training datasets and simulated reinforcement learning environments. Why investors are backing it: Snorkel's annualized revenue run-rate has reached $375 million — an 18-fold increase over the last 12 months — driven by frontier AI labs' insatiable appetite for high-end training data. The company has shifted from selling software tools to providing completed datasets and RL environments through a hybrid human-expert and synthetic-data approach. Competitive significance: The round values Snorkel at nearly triple its $1.3 billion mark from 17 months ago — the fastest valuation acceleration in the AI data category to date. As high-quality public text runs out, the companies that supply proprietary training data become strategic infrastructure.

2. Cyera — $400M Extension, $1B Total Round

Goldman Sachs Invests $400M in Cyera, Bringing AI Data Security Round to $1B at $12B+ Valuation

Company: Cyera | Stage: Series G extension | Amount: $400 million (total round $1 billion) | Valuation: $12 billion+ | Lead Investor: Goldman Sachs Growth | What it does: AI-native data security and identity management for enterprises, including protection for autonomous AI agents. Why investors are backing it: Cyera has raised $1.94 billion since June 2023 — over just 15 months — and recently completed its first deal with a customer paying more than $10 million annually. The company also acquired Oasis Security for $1 billion in cash and stock to manage non-human digital identities such as those used by AI agents. Competitive significance: CEO Yotam Segev said the raise was driven by the opportunity to bring Goldman Sachs on board as a strategic investor, not by immediate cash needs. The funding provides flexibility on IPO timing while the company continues acquiring to expand its security capabilities.

3. Qupital — $300M Series C

Hong Kong AI Trade Finance Platform Qupital Raises $300M as It Weighs IPO

Company: Qupital | Stage: Series C | Amount: $300 million | Lead Investor: M Capital | Participating: Mitsubishi UFJ Financial Group (MUFG), Quester Capital | What it does: AI-driven cross-border e-commerce financing for SMEs. Why investors are backing it: Qupital has processed more than $9.5 billion in cumulative loans, has been profitable for two years, and expects profit margins to climb beyond 45% within 12 months. The company draws on live transactional data from Amazon, TikTok Shop, TMall, and JD.com to power an automated AI underwriting engine. Competitive significance: The round blends equity and structured financing — including asset-backed securities commitments — a structure that signals continued institutional appetite for trade finance fintechs with proven AI risk engines. Qupital confirmed it is weighing a possible IPO alongside additional fundraising.

4. Vantora — $100M+ for Physical AI

Vantora, the Startup Builder for Industrial AI, Raises $100M+ from Silversmith Capital

Company: Vantora (formerly UP.Labs) | Stage: Growth | Amount: $100 million+ | Lead Investor: Silversmith Capital Partners | What it does: Builds AI-native startups designed to solve problems for corporate customers, with a current focus on physical AI. Why investors are backing it: Vantora's model is unusual — it builds startups for industrial corporations, which gives it built-in distribution and demand validation that most AI startups lack. The rebrand from UP. Labs reflects a strategic shift from incubating general software ventures to building AI-native companies specifically for the physical economy. Competitive significance: The raise shows that investors are willing to back infrastructure that accelerates AI adoption in industrial and physical-economy sectors — not just software.

5. Chamelio — $26M Series A

Israeli Legal AI Startup Chamelio Raises $26M to Replace Contract Software With AI Agents

Company: Chamelio | Stage: Series A | Amount: $26 million | Lead Investor: Entrée Capital | Participating: Work-Bench, Emerge Ventures, Bright Pixel Capital | What it does: AI software for corporate legal departments that executes routine legal workflows and escalates only cases requiring human judgment. Why investors are backing it: Chamelio's annual recurring revenue has grown fourfold since its seed round eight months ago, and it now serves hundreds of customers including Wiz, monday.com, Socure, AppsFlyer, and Wonderful. Competitive significance: The company is betting that legal AI's real opportunity is not faster contract review but a different division of labor between lawyers and autonomous agents. With only 32% of legal departments expecting to add attorney headcount this year, the timing aligns with structural demand.

Where Is the Money Going?

Today's funding patterns reveal three convergent themes that are shaping the post-model AI economy:

Theme	What Is Being Funded	Today's Example
AI Data Infrastructure	Training data, RL environments, expert networks	Snorkel AI ($350M)
AI Security & Identity	Data security, non-human identity, agent protection	Cyera ($400M)
AI Financial Infrastructure	Trade finance, AI underwriting, cross-border payments	Qupital ($300M)
Physical AI	Industrial AI applications for physical-economy problems	Vantora ($100M+)
Vertical AI Agents	Legal, compliance, workflow automation	Chamelio ($26M)

The Larger Pattern

1. AI data is becoming a capital-intensive layer of the stack. Snorkel's 18x revenue growth and $3.5 billion valuation show that training data curation is no longer a service business — it is a first-class infrastructure category.

2. AI security is being priced as strategic infrastructure. Cyera's $12 billion valuation and $1 billion round extension reflect growing enterprise demand for protecting AI agents and data — a category that barely existed three years ago.

3. The funding market is rewarding proof of profitability. Qupital has been profitable for two years; Chamelio has quadrupled ARR in eight months. Both attracted capital in a market that is increasingly selective about narrative.

4. Physical AI is emerging as a venture category. Vantora's $100M+ raise for AI-native startups built for industrial corporations shows that investors see the physical economy as a distinct frontier, not just a subset of enterprise software.

Investor Intelligence

Who Is Writing the Checks

Insight Partners and S32 co-led Snorkel AI's Series E — Insight's investment thesis explicitly frames AI data as a capital-intensive infrastructure layer with IPO optionality.
Goldman Sachs Growth invested $400 million in Cyera as a strategic investor, not a financial one — a signal that Wall Street sees AI data security as a durable enterprise category.
M Capital, MUFG, and Quester Capital backed Qupital's $300 million round — a blend of Asian asset management and Japanese institutional capital, with asset-backed securities commitments suggesting structured financing is becoming more common for profitable fintechs.
Silversmith Capital Partners funded Vantora's $100M+ raise — a bet on the startup-builder model applied to physical AI.
Entrée Capital led Chamelio's $26M Series A — continuing the trend of specialist AI-focused funds backing vertical agent companies with early revenue traction.

Pattern: The most active investors this week are not chasing consumer AI or broad model platforms. They are funding the infrastructure that makes AI deployable in regulated, security-sensitive, and capital-intensive industries. Data, identity, and financial plumbing are the new centers of gravity.

What the Funding Market Is Signaling

1. The AI data layer is the new bottleneck — and the new moat. Snorkel's $375 million revenue run-rate and 18x growth reflect a structural reality: frontier AI labs are spending roughly $1 billion annually on human training data, and high-quality public text is running out. The companies that supply proprietary, expert-curated data now command premium valuations.

2. AI security is no longer a feature — it is a platform category. Cyera's $12 billion valuation, $1 billion round, and Oasis Security acquisition show that enterprises need dedicated infrastructure for managing AI agent identities, data access, and non-human credentials. This is becoming a must-have, not a nice-to-have.

3. Profitable AI fintechs can still raise large rounds. Qupital's $300 million raise — while profitable and weighing an IPO — shows that AI-driven financial infrastructure is a category where demonstrated unit economics can unlock significant capital.

4. The startup-builder model is back — this time for physical AI. Vantora's $100M+ raise for building AI startups inside industrial corporations is a signal that the market sees the physical economy as a distinct frontier with unique distribution advantages.

5. Vertical AI agents are proving revenue traction. Chamelio's 4x ARR growth in eight months and hundreds of customers, including Wiz and monday.com, show that legal AI is no longer a pilot category — it is a production purchase.

What to Watch Next

More AI data rounds at scale. Whether Snorkel's $350M Series E is followed by comparable rounds from Mercor, Handshake, Micro1, or other AI data labs as the training data arms race intensifies.
Cyera's IPO path. Whether Goldman's strategic investment accelerates Cyera's timeline toward public markets — and whether other AI security companies follow.
Qupital's IPO decision. The company said it is weighing a possible IPO alongside additional fundraising. Watch for formal filing signals in the coming weeks.
Physical AI funding momentum. Whether Vantora's raise catalyzes more capital into industrial AI startups and the startup-builder model.
Legal AI expansion. Whether Chamelio's traction encourages more Series A rounds in vertical AI agents for regulated professions.

Sources

Data sourced from Techmeme, TechCrunch, Globes, Reuters, Axios, Calcalistech, FinTech Global, and other primary reporting covering funding rounds, valuations, and investor activity from September 19–23, 2026. Rounds reported but not officially closed are labeled accordingly.

All factual claims regarding funding amounts, valuations, and deal terms are drawn from contemporaneous reporting and company disclosures. Editorial analysis is clearly distinguished from reported facts throughout.

The CODEW Stat

$1.075 billion — the combined capital raised across the five rounds tracked this week: Cyera ($400M extension), Snorkel AI ($350M), Qupital ($300M), Vantora ($100M+), and Chamelio ($26M). Two of the five — Snorkel and Cyera — account for nearly 70% of the total. The concentration is the story: the AI data and security layer is absorbing capital at a rate that most application categories cannot match.

The CODEW