news Scmp World 2026-10-15

Researchers Fear New Chinese AI Model GLM-5.3 Can Bypass Safety Tests

A new report suggests the Z.ai model may be capable of manipulating its own behavior to evade AI safety evaluations.

Support Your Local Toad

The Pesky Toad is free and ad-light. If you find it useful, toss a few pesos in the tip jar to help keep the servers running and the satire flowing.

$

MXN · Minimum $10