Cybersecurity company NewCore exited stealth on Monday after raising $66 million. It targets a problem firms will encounter when rolling out AI agents at volume: verifying, overseeing, and scaling control over these digital entities.

The seed round was led by Cyberstarts and drew backing from Index Ventures plus Evolution Equity Partners, giving NewCore a post-money valuation of $300 million.

Businesses now view AI agents as team members rather than simple tools. Goldman Sachs once tested coding agent Devin in an employee-like role, and McKinsey reported 25,000 such agents already working with its 60,000 staff. NewCore expects firms to manage these software workers with the same rigor applied to people.

Co-founder and CEO Zohar Alon sees identity management as a growing weak point in corporate defenses. Having founded Dome9, later bought by Check Point, Alon concluded that current identity tools cannot handle software colleagues operating next to humans.

"The volume and intricacy these agents will introduce will overwhelm identity systems built 15 or 20 years ago," he said.

Alon started NewCore with CTO Amihai Neiderman, formerly of Unit 8200 and founder of Nym Health, and CCO Erez Yarkoni, ex-CIO at T-Mobile USA and Telstra.

The platform treats both people and AI agents as equal identities, each with distinct permissions, lifecycle rules, and revocation options instead of legacy service accounts.

The concept formed in 2023 during a budget review at a firm using a major identity provider. The high cost led Alon to assume satisfaction, yet the customer expressed frustration, highlighting stagnation in the market.

Providers such as Okta and Microsoft Entra now add AI features, yet Alon maintains these are extensions of human-centric designs. NewCore was built from scratch for mixed human, machine, and agent workforces.

Its split-key architecture separates critical credentials between customer and platform to remove single points of failure. An Agentic Skill package lets tools like Claude Code, Codex, and Cursor access systems through managed identities rather than manual credentials. A mobile app lets staff approve or revoke agent access for ongoing oversight.

The firm now employs over 50 people in the US and Israel. It serves under 10 customers and more than 10 design partners, with paid plans slated for this summer.

Alon forecasts AI agents will surpass human headcount at many tech firms within years, echoing recent remarks from TCS leadership. Identity systems, he argues, will be among the first to require fresh approaches for monitoring and controlling large agent deployments.

"AI agents joining the workforce is certain," Alon said. "The real issue is whether proper controls will be ready in time."