NewCore builds enterprise identity software that lets companies manage human workers and AI agents in one system. On June 15, 2026, the stealth startup said it raised $66 million in seed funding to tackle a problem a lot of companies are about to hit: once AI agents start touching production systems, old identity stacks start to look shaky. Co-founder and CEO Zohar Alon started the company with CTO Amihai Neiderman and CCO Erez Yarkoni after the idea took shape in 2023, when Alon saw customers paying huge identity bills without much real satisfaction. Cyberstarts led the round, with Index Ventures and Evolution Equity Partners joining, and it values NewCore at $300 million after the investment.
What is NewCore's AI agent identity platform and how does it work?
Giving AI agents their own identities
Here’s the plain-English version: NewCore gives every AI agent its own governed identity. It then sits inline on the identity layer so every authentication request, token, and authorization decision runs through the platform in real time. That means a company can let an agent sign into enterprise systems, but only with short-lived access tied to a specific task instead of broad standing privileges. It can run as a standalone system or alongside an existing identity provider. That matters because most enterprises won’t rip out Okta or Entra overnight.
Agentic SSO and task-scoped access
The workflow is more concrete than the usual “AI security” hand-waving. NewCore’s Agentic SSO gives agents an auditable sign-in path to connected tools. Its task-scoped tokens are minted at the narrowest permission level the destination app allows, so an agent gets access for the job in front of it, not a permanent all-you-can-eat credential. The platform also keeps an AI inventory and audit trail. Security teams can see which human or agent touched what, and when.
Managing AI agents across enterprise systems
That setup targets a pretty obvious mess inside enterprises right now: developers and employees are wiring Claude Code, Codex, Cursor, and internal copilots into production systems with whatever credentials are handy. NewCore’s “Agentic Skill” package is meant to clean that up by letting those coding agents access enterprise systems as managed identities instead of borrowed secrets or manually distributed keys. For customers, the before-and-after is simple. Before, AI tools piggyback on service accounts. After, they show up as first-class identities with permissions and lifecycle controls. They also get revocation paths of their own.
Security features and human oversight
NewCore is also trying to make the underlying identity plumbing harder to break, not just easier to administer. Its Secure Split Key model divides signing authority between the vendor and the customer environment so neither side can sign alone. The company also layers in VisualMFA and hardware-bound credentials anchored in TPM or Secure Enclave for human users. It also offers a mobile app that lets employees grant, review, or yank access for agents when human oversight is needed. That’s not a small detail. It’s the difference between “we deployed agents” and “we know how to stop them.”
Who founded NewCore and why now?
Alon’s story is the core of this company.
The founding story
NewCore started with a blunt observation. In 2023, while helping review one company’s tech budget, Alon saw what it was paying a major identity provider and assumed the customer must be thrilled. He asked, “You must be extremely happy with them.” The answer was: “No, I’m not.” That exchange pushed him toward a market he saw as big, expensive, and sleepy — right as AI agents were starting to move from demos into actual work. Alon has said the rise of software workers convinced the founders that 15- to 20-year-old identity platforms would crack under the scale and complexity ahead.
Why this team fits the market
NewCore’s founding bench is unusually strong for a seed-stage security company. Alon previously founded Dome9, the cloud-security startup Check Point acquired in 2018, and earlier in his career he helped build Provider-1, one of Check Point’s foundational enterprise products. Amihai Neiderman, NewCore’s CTO, led research in Israel’s Unit 8200 and previously founded healthcare AI startup Nym Health. Erez Yarkoni, the commercial co-founder, brings operator credibility from the buyer side after serving as CIO at T-Mobile USA and Telstra.
That mix matters. A lot of AI-security startups have strong researchers but weak enterprise selling muscle, or the reverse. NewCore has a founder who has already sold into security teams. It has a CTO who has built applied AI systems. And it has a third co-founder who has lived through big-company identity and IT pain from inside the building. That’s one reason investors were willing to back a seed round this large.
Early traction and the $66M seed round
The company has more than 50 employees across the U.S. and Israel, or more specifically Tel Aviv and the U.S. on the company’s launch materials. It’s still early: NewCore has fewer than 10 customers using the platform and more than 10 design partners, and it expects to start charging in summer 2026. That’s not massive commercial traction yet. But it’s enough to show this isn’t just a concept deck with an AI label slapped on top.
The financing is big by any seed-round standard. Cyberstarts led the $66 million round, with Index Ventures and Evolution Equity Partners participating, and the post-money valuation lands at $300 million. For a company just emerging from stealth, that price tells you investors think identity for AI agents could become its own control plane category — not just a feature inside somebody else’s admin console.
How NewCore compares with Okta, Microsoft Entra, and Aembit
This is where the pitch gets interesting. Okta and Microsoft Entra are both adding tools for AI agents and non-human identities, which means the incumbents clearly see the same opening. Microsoft now offers Entra Agent ID and governance workflows for agent identities. Okta has rolled out “Okta for AI Agents” and a broader secure-agentic-enterprise blueprint. So NewCore isn’t inventing the problem. It’s racing bigger companies that already own enterprise identity budgets.
NewCore’s counterargument is that the incumbents are bolting agent controls onto platforms built for browser logins, SAML-era federation, service accounts, and static credentials. Alon’s line is that those systems were designed for human employees first, while NewCore was built from scratch for a workforce of humans, machines, and agents. The split-key architecture and inline token control are part of that case. So are agent lifecycle management and the side-by-side deployment model.
There’s also a younger-company comparison. Aembit has been pushing IAM for agentic AI and broader workload identity, so it’s one of the clearer startup peers in this category. But Aembit’s roots are closer to workload and machine access management, while NewCore is pitching itself as a unified workforce identity layer where humans and agents live in the same plane. That distinction could matter if buyers want one policy model for both employees and software workers. Or it could blur fast if the category gets crowded.
Why this AI agent identity round matters
Because this round isn’t really about selling another dashboard.
It’s about giving NewCore enough capital to attack a category that usually favors giants. Identity is sticky, high-risk software. Buyers hate replacing it. So if a startup wants to break in, it needs enough engineering muscle to ship core security architecture. It also needs enough product depth to coexist with legacy systems, and enough go-to-market patience to win cautious enterprise customers. A $66 million seed round gives NewCore room to do that before revenue fully ramps.
It also signals that investors believe AI-agent governance won’t stay a side module forever. If enterprises really do move from a few copilots to fleets of semi-autonomous agents, the company controlling identity, authorization, and revocation becomes extremely valuable. That’s the thesis here. Alon put it more bluntly: “It’s inevitable.” The real question, he said, is whether companies build those guardrails in time.
Why is AI agent identity becoming a real market?
The macro numbers are starting to line up. Gartner said in June 2025 that 24% of CIOs and IT leaders in a webinar poll had already deployed at least a few AI agents, and another 4% had deployed more than a dozen. Separate market research put the broader IAM market at $26.77 billion in 2025, with a forecast of $62.90 billion by 2033. Even if those forecasts move around, the direction is obvious: identity is getting bigger, not smaller, and AI is dragging non-human access into the center of that spend.
The enterprise anecdotes are getting hard to ignore. Goldman Sachs tested the AI coding agent Devin as a new employee in 2025. McKinsey said earlier in 2026 that 25,000 AI agents were already working alongside its 60,000 employees. TCS chairman N. Chandrasekaran has made a similar point, saying AI agents could eventually rival his company’s workforce in size. If that’s even half right, identity teams aren’t heading toward a mild upgrade cycle. They’re heading toward a scale problem.
That’s why NewCore’s timing makes sense, even if the company still has a lot to prove. Legacy identity tools were built for people logging into apps. Agentic systems create nonstop token requests, sub-agents, delegated actions, and a flood of machine-speed decisions. NewCore’s own view is dramatic — identities could grow roughly 100x and identity events 100x as agents mature — but even a softer version of that claim points to the same thing: this category is no longer theoretical.
Can NewCore win the AI agent identity race?
Maybe. But it won’t win just by naming the problem first.
NewCore has a serious founding team and a very large seed round. It also has a product story that’s more specific than most AI-security launches. Now it has to beat timing risk, prove that enterprises want a new identity layer instead of extensions from Okta or Microsoft, and convert design partners into paying customers starting in summer 2026. That’s what to watch next.
Read how AutoVRse raised $2.4M from Singularity AMC and Lumikai to expand its enterprise VR training platform that helps industrial companies turn frontline expertise into scalable, AI-powered immersive learning workflows.
FAQ
- What is the NewCore funding round? NewCore raised a $66 million seed round announced on June 15, 2026. Cyberstarts led the financing, with Index Ventures and Evolution Equity Partners participating, and the deal valued the company at $300 million after the investment.
- How does NewCore’s platform work for AI agents? It gives each AI agent its own enterprise identity instead of treating it like a shared service account. The platform handles sign-in and issues short-lived task-scoped tokens. It evaluates access requests in real time and keeps an audit trail so companies can monitor and revoke agent access when needed.
- Who founded NewCore? NewCore was founded by Zohar Alon, Amihai Neiderman, and Erez Yarkoni. Alon previously founded Dome9, Neiderman led research in Unit 8200 and founded Nym Health, and Yarkoni was CIO at both T-Mobile USA and Telstra — which gives the team a mix of security, AI, and enterprise IT experience.
- Is AI agent identity really a separate market category? It’s starting to look like one. Big incumbents like Microsoft and Okta now ship agent identity features. Newer vendors like Aembit are targeting agentic AI access. Gartner’s 2025 poll showed that a meaningful slice of enterprise tech leaders had already deployed at least some AI agents.
