NewCore Raises $66M to Build AI-Agent Identity Management Platform, Valued at $300M
Summary
- • NewCore raises $66M seed round led by Cyberstarts to give AI agents proper digital identities within enterprise systems.
- • The startup is valued at $300M and targets the gap left by legacy identity platforms like Okta and Microsoft Entra.
- • Its platform unifies human and AI-agent identity management with a 'split-key' credential architecture to eliminate single points of compromise.
- • Integrations with Claude Code, OpenAI Codex, and Cursor allow AI coding tools to authenticate as managed enterprise identities.
Details
NewCore raises $66M seed round at a $300M post-money valuation
Round led by Cyberstarts; Index Ventures and Evolution Equity Partners also participated.
Platform manages human and AI-agent identities in a single unified system
AI agents are treated as first-class identities with their own permissions, lifecycle controls, and revocation mechanisms — not traditional service accounts.
Split-key credential architecture eliminates any single point of compromise
Critical identity credentials are divided between the customer and the NewCore platform, preventing a central breach from exposing all agent access.
Agentic Skill packages integrate with Claude Code, OpenAI Codex, and Cursor
Allows AI coding assistants to access enterprise systems as managed identities rather than through manually distributed credentials.
Mobile app gives employees real-time control over AI agent access permissions
Employees can grant, review, and revoke agent access on demand — providing a human oversight layer as companies deploy more autonomous systems.
McKinsey already has 25,000 AI agents working alongside 60,000 human employees
Goldman Sachs also tested coding agent Devin as a new employee — illustrating the enterprise scale that drives NewCore's market thesis.
NewCore expects to begin charging customers in summer 2026 with fewer than 10 paying clients today
Currently serving more than 10 design partners; commercial launch this summer will be an early test of enterprise willingness to pay for dedicated AI-agent identity.
NewCore emerges from stealth to address AI-agent identity governance as enterprises deploy autonomous agents at scale.
What This Means
NewCore's $300M valuation at seed stage signals that AI-agent identity management is rapidly becoming a serious enterprise security category. As AI agents gain autonomous access to corporate systems, the inability of legacy identity platforms to govern them creates significant and growing security risks. The integration with major coding assistants — Claude Code, Codex, Cursor — positions NewCore at the developer layer first, a classic enterprise security go-to-market strategy. Incumbents like Okta and Microsoft are retrofitting AI-agent features onto human-identity foundations, giving purpose-built startups like NewCore a window to establish themselves before the market consolidates.
Sentiment
Mostly intrigued, highlighting the emerging need for AI agent security infrastructure
“A startup just came out of stealth with $66M because AI agents are about to start logging into work. NewCore is building identity security for companies where humans, machines, and agents all need access to internal systems. It is already valued at $300M... Imagine a coding agent inside Claude Code or Cursor. It needs GitHub access... Every useful agent needs a badge.”
“UPDATE: Startup NewCore has raised $66 million to help companies manage AI agents like employees, giving them permissions, access controls, and the ability to be monitored or shut off.”
“NewCore raised $66M to give AI agents identities. We went from 'agents will replace employees' to 'agents need IAM and SSO' real quick. Plumbing eats the narrative every time.”
“AI agents are becoming employees. Someone's building them ID cards. @TechCrunch reports NewCore raised $66M for agent identity. Okta for bots. When agents touch real workflows, identity isn't theoretical. Tracking which agents do what?”
Split
~80/20 positive/neutral — broad agreement on the security gap, minimal pushback or skepticism in visible discussion.