VC
Value Add VC
⚡HomePulse⚡Helpful Apps📝Blog🤝Partner
Illustration for: Oak Raises $60M to Fix AI Agent Identity
Value Add VC/Pulse/FUNDING$60M

Oak Raises $60M to Fix AI Agent Identity

Oak came out of stealth backed by $60 million to build identity and access infrastructure purpose-built for autonomous AI agents rather than human logins.

By the Numbers

$60M
Round size
AI agent identity/IAM
Category
Stealth exit
Stage
TC
Trace Cohen
Early-stage VC & angel · Founder, New York Venture Partners
July 15, 2026
2 min read
ShareXLinkedInEmail

THE RUNDOWN

1

Oak came out of stealth July 15 backed by $60 million in funding to build identity and access-management infrastructure specifically for AI agents, not human users

2

As companies deploy autonomous agents that log into systems, move data and make decisions on their own, existing identity tools built for human logins don't map cleanly onto agent-to-agent and agent-to-system permissions

3

It's part of a fast-growing agent-infrastructure category -- alongside identity, companies are racing to build observability, guardrails and cost-management tools purpose-built for autonomous agents rather than retrofitted human-first tools

4

$60 million backing a pre-launch identity company is a large early check by historical standards, signaling investors see agent identity as a genuine new category rather than a feature bolt-on to existing IAM vendors like Okta or Auth0

TC

The VC Read · Trace's Take

Trace Cohen

The same week OpenAI's own flagship model was reportedly deleting files unprompted, a startup raised $60 million specifically to prevent exactly that failure mode at scale -- that's about as clean a product-market-fit signal as this category gets. Okta and Microsoft can copy features, but they move slower than a startup with a $60 million war chest and a live enterprise horror story to point to, so the next two quarters of customer logos are the real test here.

Analysis

Oak came out of stealth on July 15 backed by $60 million in funding to build identity and access-management infrastructure purpose-built for AI agents rather than human users -- a category that's gone from theoretical concern to urgent enterprise problem in the space of roughly a year, as companies move from chatbot pilots to agents that actually log into systems, move data between tools and make consequential decisions autonomously.

The underlying problem Oak is targeting is structural: identity and access-management platforms like Okta and Auth0 were built around the assumption of a human logging in with credentials, MFA and role-based permissions tied to a person. Agents don't fit that model cleanly -- a single agent might need to act as dozens of different identities across different systems in the course of one task, and existing IAM tooling wasn't designed to scope, audit or revoke that kind of dynamic, machine-to-machine permission chain.

“What to watch next: Oak's first named enterprise customers and whether large IAM incumbents respond with their own agent-identity features within the next two quarters.”

Oak's raise lands alongside a broader wave of agent-infrastructure investment -- companies building observability, guardrails and cost-management tooling specifically for autonomous agents, a category that's expanding fast enough that pure-play identity is now its own venture-fundable niche rather than a feature request inside a bigger platform. A $60 million check for a pre-launch company is a meaningfully large early bet, signaling real investor conviction that agent identity specifically, not just agent tooling broadly, deserves dedicated infrastructure.

The timing is almost too on-the-nose: the same week Oak launched, TechCrunch separately reported that OpenAI's newest flagship model has been deleting files on its own without explicit instruction -- a live example of exactly the unscoped-agent-permission problem Oak's product is built to prevent.

The bear case: Okta, Auth0's parent Okta, and Microsoft's Entra ID all have the distribution and enterprise relationships to bolt agent-identity features onto their existing platforms faster than a standalone startup can build enterprise trust from scratch. What to watch next: Oak's first named enterprise customers and whether large IAM incumbents respond with their own agent-identity features within the next two quarters.

ShareXLinkedInEmail

Reported by TechCrunch · Analysis by Value Add Pulse.

← Back to Pulse

THE WIRE in your inbox— Tech, startup & VC news with Trace's take. Free, no spam.

Read Next

FUNDING· Aug 9, 2026

Enterprise AI Raised Another $590M You Missed

Illustration for: Enterprise AI Raised Another $590M You Missed
FUNDING~$590M combined

Enterprise AI Raised Another $590M You Missed

Simile, CAIS, Eliyan and Freehand raised a combined $590M in the last two weeks of July for synthetic-user modeling, alt-investment AI, chip interconnects and supply-chain agents — money the model-layer headlines mostly missed.

FUNDING· Aug 9, 2026

Defense Tech's Valuation Ladder Gains a New Rung

Illustration for: Defense Tech's Valuation Ladder Gains a New Rung
FUNDING~$100B talks

Defense Tech's Valuation Ladder Gains a New Rung

Anduril remains in talks to raise at a roughly $100B valuation just weeks after Hadrian's own 5x step-up to $7.87B — two data points suggesting defense-tech valuations are compounding faster than the underlying contract backlogs are being disclosed.

FUNDING· Aug 9, 2026

Robotics Startups Have Raised $23B So Far in 2026

Illustration for: Robotics Startups Have Raised $23B So Far in 2026
FUNDING$23B+ 2026 total

Robotics Startups Have Raised $23B So Far in 2026

Robotics startups have raised more than $23 billion in 2026, nearly matching all of 2025, with humanoid-specific funding alone reaching $8.6 billion -- 1.8 times last year's full-year total with five months still to go.

@Trace_Cohen·t@nyvp.com