Cygrid exits stealth with €2.66M to enforce what AI agents do after sign-in

Cygrid (Berlin) has exited stealth with a €2.66 million Pre-Seed round led by Atlantic, the pan-European pre-Seed fund, with participation from D11Z.Ventures and a group of angels including Boris Gorin (VP Product at Zscaler, founder of Canonic Security), Andreas Bermig (CEO of bonify), and Vlad Tropko of AmberVale Capital. The company was founded in January 2026 by Dmitry Kharchenko (CEO) and Dekel Paz, both veterans of Israeli Military Intelligence. In nine months, it has signed 19 design partners — six already in production.

The problem: access control stops at the door

Conventional identity and access management answers one question: can this identity authenticate? Once the answer is yes, IAM's job is done. What happens next — what actions the identity actually takes, on whose behalf, against which systems, with what data — is largely unmonitored.

That gap was tolerable when the identities in question were humans with finite working hours. It becomes consequential when the identities include AI agents executing at machine speed, service accounts held by third-party applications, and non-human pipeline actors that may have acquired tokens long after their original purpose. Palo Alto Networks places the current ratio of machine to human identities at 109:1; 79 of those machine identities are now AI agents. CrowdStrike's 2026 report found that 82% of detections involve no malware at all — the attack vector is a trusted identity that has been leaked, misused, or never properly scoped.

Cygrid's response is continuous enforcement: a layer that sits above the authentication stack and applies real-time controls to every action an agent or identity attempts to take, not just every log-in. The platform maps human, non-human, and AI agent identities into a unified intelligence graph — what each identity is entitled to do, what it actually does, and on whose behalf it acts. A forgotten token can be revoked in seconds; a risky AI agent command is blocked before it executes.

In one early deployment, a Berlin-based D2C retailer discovered AI agents and third-party application connections its own security team was unaware of, and used Cygrid to remediate more than 100 risky findings.

The investor case

"Dmitry and Dekel are proven builders who combine deep cybersecurity experience with a real understanding of what enterprises need: security that stops the wrong action without slowing the business down," said Marc-Olivier Lücke, General Partner at Atlantic. "AI agents are widening the identity attack surface faster than any control can keep up. Intelligence-driven enforcement is the only answer that scales."

The angel syndicate carries operational weight. Gorin built Canonic Security — which addressed similar non-human identity risks before acquisition — and is now VP Product at Zscaler, which has a financial stake in the category expanding. Bermig runs bonify, a German consumer finance platform that has grappled with non-human identity proliferation firsthand. The angels are not decorating a cap table; they are writing from experience of the problem.

Cygrid will use the €2.66M to expand its enforcement capabilities, grow its technical team, and scale go-to-market across Europe. The stated ambition is to become the European alternative to US and Israeli AI identity security vendors — a sovereignty argument that is becoming commercially relevant under the NIS2 and DORA regulatory frameworks.