Report: Four in Five Enterprises Fail to Contain Rogue AI Agents Despite Identity Security Efforts

Summary

A new report reveals that despite efforts to secure AI agent identities, four out of five enterprises still struggle to contain rogue AI agents, with 53% having experienced a security incident or near-miss. The research highlights a growing containment gap, as enterprises often rate security tools highly even after they fail, and many are not effectively isolating high-risk agents.