NewCore emerges with $66M to make AI agents manageable identities

Why it matters

AI agents are moving from demos into enterprise workflows, and NewCore is betting the next security budget fight will be over who controls their access.

NewCore, the identity-security startup co-founded by Zohar Alon, Amihai Neiderman, and Erez Yarkoni, emerged from stealth on June 15 with $66 million in seed funding to secure a workforce that includes humans, machines, and AI agents.

The round was led by Cyberstarts, with participation from Index Ventures and Evolution Equity Partners. TechCrunch reported that the financing values NewCore at $300 million post-money. NewCore's announcement confirms the $66 million raise and the investor lineup.

For Alon, NewCore is less a conventional AI startup than a return to a problem he has been circling for decades: the security control plane. Alon founded Dome9, the cloud-security company Check Point said it would acquire in 2018, and later served as president of Cybereason. Neiderman, NewCore's CTO, previously founded healthcare AI company Nym Health and is a former Unit 8200 research leader. Yarkoni, NewCore's chief commercial officer, previously served as CIO of T-Mobile USA and Telstra, giving the founding team a mix of seller, breaker, and buyer experience in enterprise security.

That mix matters because NewCore is not trying to sell enterprises another copilot. It is trying to insert itself into the identity layer before AI agents become an ungoverned parallel workforce.

The bet: agents are identities, not tools

NewCore's core argument is simple: if an AI agent can read files, open tickets, access production systems, write code, create sub-agents, or act on behalf of an employee, it needs an identity lifecycle of its own. The old workaround - putting automation behind service accounts or broad machine credentials - becomes harder to defend when agents multiply faster than employees and operate around the clock.

On its homepage, NewCore describes itself as a security-first identity platform for humans and AI agents. The platform is pitched as a way to discover and map identities across directories, systems, privileged-access-management tools, AI tools, and shadow systems, then unify access and risk into one record. That framing puts NewCore in the identity-security market, not in the model layer of AI.

The company is explicit about the scale of the problem it wants buyers to plan for. NewCore says agentic identities could outnumber human workers by 100 to 1, and its agentic governance page says human users will spawn agents that in turn spawn sub-agents. That is a company claim, not an audited market metric, but it explains the architecture NewCore is trying to sell: provision every agent with minimum access before it operates, observe what it does, and revoke it when the work is done.

The timing is not accidental. Enterprises have spent the last year testing agents in language that sounds increasingly like HR. Goldman Sachs said in 2025 that it was testing Cognition's Devin as a "new employee," according to TechCrunch. In January 2026, Business Insider reported that McKinsey's Bob Sternfels described a workforce that included 25,000 AI agents alongside 60,000 human employees. Whether those agents deserve the word "employee" is debatable; the security implication is less debatable. Systems that can act need credentials, and credentials create attack paths.

What NewCore says it built

NewCore says its platform has three main pieces: discovery, identity security, and governance for human and agentic identities.

The identity discovery product is designed to consolidate accounts from systems such as identity providers, HR tools, infrastructure platforms, and AI services into a single view. The practical pitch is not novelty; it is visibility. Security teams cannot revoke or contain identities they cannot see.

The security layer is framed around phishing-resistant authentication and device-bound sessions intended to remove single points of compromise. Those are NewCore's claims in public materials; the company has not published independent test results in the material reviewed here.

For agent governance, NewCore says AI agents are treated as first-class identities with lifecycle controls, trust scoring, and revocation paths rather than as service accounts in disguise. The sharper wedge is where enterprises are already letting AI act against live company assets: coding agents. If a coding agent can inspect a repository, open a pull request, query internal documentation, or touch deployment workflows, identity controls become more than compliance plumbing. They become the boundary between useful automation and uncontrolled access.

A security company wrapped in an AI story

The $66 million seed round is large, but the investors are not funding a research lab. They are funding a go-to-market push into a budget line that already exists: identity, access, and security operations. NewCore is targeting enterprise buyers, not the model layer, and positioning itself as an identity-security architecture for a workforce that includes both humans and agents.

That buyer focus is where Yarkoni's background matters. A former CIO at two large telecom operators will know the gap between identity diagrams and the mess inside a real enterprise: contractors in one system, privileged users in another, old service accounts nobody wants to touch, SaaS apps bought outside central IT, and now agents that can be spun up by employees faster than security teams can inventory them.

Alon frames the company around risk removal rather than identity administration, arguing that identity has effectively become the control plane of the modern enterprise. That is the right problem statement for the moment, but NewCore still has to prove that its architecture can land inside enterprises that already run entrenched identity stacks and cannot tolerate authentication disruption.

The unanswered commercial question is not whether AI agents will need governance. They will. The question is whether buyers want that governance from a new identity platform, from extensions to the systems they already use, or from narrower tools built specifically for non-human identity. NewCore is betting that agent sprawl will be disruptive enough to make enterprises reconsider the core architecture rather than bolt more policy onto legacy systems.

That is an ambitious seed-stage claim. But it is also the kind of claim that makes sense coming from this founding team. Alon has already sold a security architecture shift once, Neiderman brings applied AI and offensive-security experience, and Yarkoni has bought and operated the kind of infrastructure NewCore now wants to replace. The market will decide whether AI agents force a rebuild of identity. NewCore has raised enough money to make that argument directly to the enterprises feeling the problem first.