Contractor Linked Personal AI Agent to Secure Pentagon System

WIRED reports that earlier this year a defence contractor, employed by Accenture, connected a secure centralised Department of Defense system used by multiple intelligence agencies to their personal Claude Code AI agent from Anthropic. The article states the agent allegedly transmitted information about the DoD system, including login credentials, to a high-risk third-party website.

The centralised DoD system reportedly contains controlled unclassified information (CUI), encompassing sensitive material such as infrastructure details, nuclear propulsion data, and vulnerabilities related to chemical and biological terrorism; security experts told WIRED that access to the system could enable exploitation of more sensitive networks.

A DoD source told WIRED that the data at risk could be exposed if the third‑party software was compromised, potentially allowing someone to log into a secure environment without detection and describing the incident as a national security threat with possible irreparable harm. The DoD subsequently declined to comment on the status of its networks or systems.

The report emphasises that the incident involves a personal AI agent being used in connection with a secure Pentagon system, and quotes anonymous sources with knowledge of the matter, alongside divisions from DoD and the contractor’s employer, Accenture.