A recent study reveals that companies are granting AI agents access to vital systems at a rapid pace without implementing the necessary security measures. This lack of comprehensive identity system security puts critical platforms like Active Directory, EntraID, or Okta at risk of being exploited by attackers. The concerning trend was highlighted in a study conducted in Hoboken, N.J. on May 13, 2026. The research underscores the urgency for organizations to prioritize safeguarding these systems to prevent unauthorized access and potential compromise.