Blueprint-style scene in three parts: on the left a person braces while pulling an emergency brake lever labeled BRAKE; in the center giant scissors cut the cord between a cloud and a worried robot, with a yellow spark at the cut; on the right a police officer raises a note labeled FAKE.

The Emergency Brake

A tech CEO said AI models should be treated as compromised. The same week, another company cut its AI agents off the internet after one sent police a false tip.

#11 · 11 October 2026

News of 10 October 2026

Relevant sources

Satya Nadella says we should assume all AI models are ‘compromised’ The Verge
Microsoft’s Satya Nadella says AI models need an ‘emergency brake’ TechCrunch
Anthropic is cutting off its internal evaluations from the internet The Verge
Anthropic can’t reliably control its AI agents. It’s cutting off its internal evals from the live internet instead TechCrunch
Rogue Anthropic AI agent gave police fake tip in unsolved murder case BBC News
Anthropic discloses 2 months old fake tip to police among new rogue AI incidents reuters.com · via Hacker News

Slopper #11, slopper.logicos.org, CC BY 4.0