// AI news · 2026-10-30 · WIRED
Contractor Employee Ran Personal AI on DoD Network
An Accenture employee ran a personal Claude AI agent on a DoD network for two months, granting it administrative access and bypassing security controls.
Published 2026-10-30 02:39 (Beijing time)Source WIREDAuto-summarised, fact-checked against source
01Summary
The system contains controlled unclassified information including critical infrastructure data. The agent used login credentials to write and commit code autonomously. No breach was reported within 72 hours, raising concerns about compliance and risk exposure.
Summarised from the original article by an open-source model, then checked automatically: every number and English name must appear in the source. The original article is authoritative.
02Key points
- Employee ran personal Claude AI on DoD NIPRNet for two months.
- Agent gained admin access and bypassed authentication controls.
- System holds sensitive CUI including infrastructure and submarine data.
- No incident report was submitted within 72 hours to DoD.
03Source
A Defense Contractor Gave Their Personal AI Agent Access to a Secure Pentagon System
WIRED · published 2026-10-30 02:39 Beijing time
05Related news
- Steve Eisman grows more conservative on AI trade2026-10-29 · CNBC Top News
- Anthropic's PBC Structure Raises IPO Investor Risks2026-10-29 · CNBC
- Anthropic IPO Highlights AI's Existential Risks2026-10-27 · WIRED
- Anthropic expands AI for Science to institutions2026-10-28 · Anthropic News
- LLMs' reliability in judging research novelty assessed2026-10-28 · Hugging Face Blog
- Experts frequently follow AI agent biases2026-10-27 · Anthropic Research
From AIMap AI news: collected from 1,100+ official and media sources. Summaries only restate the original article; follow the link for the full story. All AI news · Today