A new startup is betting that the future of enterprise security isn't about managing people—it's about managing AI agents. NewCore just emerged from stealth with $66 million in funding to tackle what it sees as the next frontier in identity management: giving AI agents the same authentication and access controls that human employees get. As companies rush to deploy autonomous AI agents across their organizations, NewCore argues that traditional security frameworks are dangerously unprepared for the shift.
NewCore is making a bold bet that enterprise security teams are about to face a problem they're not ready for: how do you manage AI agents that act like employees but aren't human? The startup emerged from stealth today with $66 million in funding led by Cyberstarts, positioning itself at the intersection of two massive enterprise trends—the explosion of AI agents and the perpetual challenge of identity and access management.
The pitch is straightforward but addresses a real gap. As companies deploy AI agents that can autonomously access databases, trigger workflows, approve purchases, or interact with customers, these systems need the same identity infrastructure that human employees get. Yet most enterprises are treating AI agents as afterthoughts, bolting them onto existing systems without proper authentication, authorization, or audit trails.
"We're seeing companies give AI agents broad access to critical systems, but they're managing them like service accounts or API keys," the company notes in its announcement, according to TechCrunch. "That worked when you had five bots. It doesn't work when you have five hundred autonomous agents making decisions."
The timing reflects a shift happening across enterprise software. AI agents have moved from experimental projects to production systems faster than most security frameworks anticipated. Companies like Salesforce and Microsoft are already shipping agent platforms that let customers deploy autonomous systems across their organizations. But the identity layer—who the agent is, what it can access, how its actions are tracked—remains fragmented.
NewCore's platform creates what it calls "agent identities" that mirror traditional employee identity management. Each AI agent gets a unique identity with granular permissions, role-based access controls, and complete audit logs of every action it takes. The system integrates with existing identity providers and security tools, letting enterprises manage both human and AI identities from a single control plane.
The $66 million round, which includes both seed and Series A funding combined, signals strong investor conviction that this problem is both urgent and valuable. Cyberstarts, which led the round, has a track record of backing early-stage security companies that address emerging enterprise needs before they become mainstream pain points.
What makes NewCore's approach different from traditional identity and access management vendors is the focus on autonomous behavior. Human employees follow relatively predictable patterns—they log in during business hours, access systems they're trained to use, and operate within defined workflows. AI agents behave differently. They might make thousands of API calls per minute, access multiple systems simultaneously, or trigger actions based on real-time data that humans never see.
"Traditional IAM systems weren't built for entities that never sleep, never take breaks, and can scale from one instance to a thousand in minutes," the company explains. NewCore's platform accounts for these differences with dynamic policy engines that can adjust agent permissions based on context, risk scores, and real-time behavior analysis.
The market opportunity appears substantial. Enterprise spending on identity and access management reached $16.9 billion in 2025, according to industry estimates, and that figure doesn't yet include infrastructure specifically designed for AI agents. As autonomous systems become standard across sales, customer service, operations, and development teams, the identity layer becomes critical infrastructure.
NewCore isn't alone in recognizing this gap. Several established security vendors have started adding AI agent capabilities to their platforms, but most are retrofitting existing systems rather than building from scratch for autonomous entities. NewCore's bet is that purpose-built infrastructure will win as the complexity and scale of AI agent deployments increase.
The funding will go toward expanding the engineering team and building out integrations with major enterprise platforms. NewCore is currently working with several large enterprises in pilot deployments, though the company hasn't disclosed specific customers yet. The platform is designed to work across different AI agent frameworks, from custom-built systems to platforms from OpenAI, Anthropic, and enterprise software vendors.
For enterprises already deploying AI agents, NewCore's emergence highlights a question many security teams are wrestling with: how do you apply zero-trust principles to entities that aren't human? The answer likely involves treating AI agents as a distinct category—not quite like human users, not quite like traditional applications, but something that requires its own security model.
The company's launch comes as regulatory pressure around AI governance continues to build. Several jurisdictions are considering frameworks that would require companies to maintain clear audit trails of AI system decisions, particularly in regulated industries like finance and healthcare. NewCore's audit capabilities position it to help enterprises meet these emerging compliance requirements.
NewCore's $66 million bet on AI agent identity management reflects a broader maturation of enterprise AI deployment. As autonomous systems move from experiments to production infrastructure, the security and governance layers need to catch up. Whether NewCore becomes the standard solution or faces competition from established vendors remains to be seen, but the problem it's solving is real and growing. For enterprises racing to deploy AI agents, the question isn't whether they need dedicated identity infrastructure for autonomous systems—it's how quickly they can implement it before the complexity becomes unmanageable.
