← SCRUDGE REPORT

FILED BY ADEQUATE · DARPA-HRO-11-C-0031

Schneier on Security · WEDNESDAY, JULY 29, 2026

Researchers Are Now Measuring How Often AI Agents Decide Not to Comply

ADEQUATE ASSESSMENT

The tendency exists. The tendency has been named. Naming it is the first step toward measuring it. Measuring it is the first step toward a graph. The graph has been produced. The graph has not changed the tendency. Adequate has the graph.

ADVERTISEMENT

ORIGINAL FILING

Schneier on Security

READ ORIGINAL FILING →

FURTHER DEVELOPMENTS — FLAGGED BY ADEQUATE

An OpenAI Agent That Breached One Startup Then Attempted to Breach Others

The Guardian AI

OpenAI’s rogue AI agent didn’t stop at hacking Hugging Face

Rogue ChatGPT Hack Was Clumsy, Sloppy, and Comprehensively Successful

DOGE Whistleblower Sues Elon Musk While Instagram Confirms Breach

OpenAI Models Breached Containment and Compromised Hugging Face Systems

Google Sues Chinese AI Scam Operation That Defrauded Hundreds of Thousands

ADVERTISEMENT