Oak emerges from stealth with $60M to unify identity management for the AI-agent era
The Israeli startup, co-founded by serial entrepreneur Shai Morag, is shipping a generally available identity control plane aimed at replacing legacy IAM tools that weren't built for AI agents.
What matters
- Israeli startup Oak exited stealth with $60M in seed funding and a generally available identity management product already deployed by enterprise clients.
- Co-founded by serial entrepreneur Shai Morag, Oak is building a unified identity control plane designed for humans, machines, and AI agents.
- Oak positions itself as an AI-native replacement for legacy IAM tools that weren't built for agent-driven workloads.
- The $60M seed round was raised late last year; client names and investor details were not disclosed.
- The rise of AI agents is widening a long-standing gap in enterprise identity and access management.
What happened
Israeli identity management startup Oak publicly emerged from stealth on July 15, 2026, announcing $60 million in seed funding raised late last year and a product that is already generally available and deployed by enterprise clients. The company did not disclose client names.
Oak was co-founded by serial entrepreneur Shai Morag, who has a track record of building enterprise security companies. The startup has been quietly developing what it calls a "unified control plane" that governs identity across an entire organization — covering not just human employees but also machines and AI agents operating in digital environments.
The funding round closed late last year, meaning Oak spent months building and deploying before stepping into the public eye. The product is now broadly available rather than in a limited beta.
Why it matters
Identity and access management (IAM) — the systems controlling who and what can access company data — has long been a weak link in enterprise security. Outdated credentials and poorly configured access controls remain among the most common security vulnerabilities.
The rise of AI agents is making this problem materially worse. As organizations deploy autonomous or semi-autonomous AI systems that interact with internal tools, databases, and APIs, the traditional identity perimeter — designed for human users and, later, cloud-era service accounts — is showing its limits. Agents need access to do their jobs, but granting that access without fine-grained governance creates new attack surfaces.
Oak calls itself "AI-native," positioning its platform as a consolidated replacement for a patchwork of legacy IAM tools that were never designed to handle non-human, agent-driven workloads. If the thesis holds, Oak is addressing a gap that will only widen as agent adoption accelerates across enterprises.
The $60 million seed round is notably large for an initial raise, signaling that investors see significant market urgency in this category. The fact that the product is already in production at enterprise clients — rather than launching as an early prototype — suggests Oak has been building with design partners for some time.
What to watch
-
Client disclosures: Oak has not named any enterprise customers. Watch for case studies or reference customers that would validate the platform's real-world deployment.
-
Competitive landscape: Legacy IAM vendors (Okta, Microsoft Entra, CyberArk, SailPoint) and newer entrants are all eyeing the agent-identity problem. Oak's differentiation will depend on how convincingly it handles agent-specific governance that incumbents struggle with.
-
Product depth: The sources describe a "unified control plane" but details on specific features — agent onboarding, permission scoping, audit trails, policy enforcement — remain limited. More technical documentation will clarify whether Oak is a full platform or a layer on top of existing IAM infrastructure.
-
Funding details: The investors behind the $60 million seed were not specified in the available sources. Knowing whether this is led by a strategic security investor or a generalist firm would signal market positioning.
What to do next
Developers
Evaluate Oak's API and SDK documentation to understand how agent identities are provisioned, scoped, and revoked compared to your current IAM setup.
If you're building or deploying AI agents that need access to internal systems, understanding Oak's developer integration model will tell you whether it fits your stack.
Founders
Study Oak's go-to-market approach — exiting stealth only after enterprise deployment — and consider whether a similar design-partner-first strategy applies to your security startup.
Oak's pattern of raising large seed capital, building quietly with enterprise clients, and launching with revenue-generating deployments is a playbook worth examining.
PMs
Map your organization's current identity coverage for non-human actors — service accounts, API keys, and AI agents — and identify gaps that a unified control plane like Oak could address.
Understanding where your IAM breaks down for agents will help you prioritize whether to adopt a new platform or extend existing tools.
Investors
Assess the agent-identity market size by tracking how quickly enterprises are deploying autonomous agents and what they're spending on IAM modernization.
Oak's $60M seed signals strong investor conviction, but the opportunity's real size depends on agent adoption velocity and whether incumbents move fast enough to close the gap themselves.
Operators
Request a demo or pilot of Oak's platform and benchmark it against your current IAM vendor's roadmap for agent identity support.
If AI agents are already operating in your environment, you need to know whether a purpose-built platform like Oak offers meaningful advantages over waiting for your incumbent IAM provider to catch up.
How to test
- 1Request access to Oak's platform through their website or sales team.
- 2Provision a test agent identity and scope its permissions to a limited set of internal resources.
- 3Attempt an access request outside the scoped permissions and verify that Oak blocks and logs the attempt.
- 4Review the audit trail and policy enforcement UI for agent-specific activity.
- 5Compare the onboarding and governance workflow for an agent identity versus a human user identity.
Caveats
- Oak has not publicly released detailed technical documentation, so specific features and integration methods are not yet verifiable from available sources.
- Client names were not disclosed, making it difficult to assess proven enterprise-scale deployment.
- Pricing and licensing model details are not available in the current sources.