Token Security is arguing that every autonomous AI agent should be treated as an identity in its own right, and is backing that pitch with new survey data and a natural-language tool called Enzo that lets teams build custom security applications on top of live identity data.
June 2026 · Token Security
Every AI Agent Is an Identity — Most Organizations Don't Treat Them That Way
Autonomous agents create, use and rotate credentials across systems like Salesforce, Snowflake and GitHub — spawning a wave of non-human identities that hand-built security workflows can't see or govern.
The Visibility Gap — 2026 Cloud Security Alliance Study
82%
discovered AI agents unknown to their security, IT or governance teams in the past year
41%
found unknown agents more than once
65%
had a security incident tied to AI agents
61%
cited sensitive data exposed or mishandled
The Trajectory
Task-specific AI agents embedded in enterprise applications — Gartner projection
As agents proliferate, the gap between fast-moving autonomous software and slow security processes widens.
Enzo — AI-Native Identity Security Builder
Teams describe a problem in plain language and build tooling on top of live identity data — no extra infrastructure, with tenant isolation and audit logging.
Agent attack-path engines
Access review & enforcement
Anomaly detection
Offboarding verification
M&A identity migration
Early Signal
One early adopter uncovered and secured hundreds of unmanaged AI agent identities within weeks — surfacing risk patterns like shared credentials and reused service principals that are hard to detect manually.
The Caveat
Favorable feedback so far is limited to vendor-provided cases . Independent reviews and critical assessments are not yet available, and real-world durability at scale remains unproven.
Continue reading The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in ✓ Signed in — this article isn’t included in your current plan.Unlocking the full article…