As cited
Copy frozen at (site build).
ai security
Forgotten AI agents may still access sensitive company data
Arrakis Security's 2026 Autonomous Agents Threat Report found that 76% of enabled artificial intelligence (AI) agents had not executed in 60 days or more, yet retained active credentials and connections to company systems. These inactive AI agents retain the ability to read and write data, invoke tools, and initiate workflows, creating a persistent access pathway. The analysis examined thousands of agents across customer production environments over a 12-week period.
Why it matters: Security teams managing AI agents need to audit and revoke credentials for inactive agents immediately, as dormant software poses an overlooked lateral movement and data access risk.
- Source published
- First seen by Cybersecurity Tracker