AI Radar
Support
LiveUpdated 2026-09-20 23:15 UTC

GPT-6 Astra attempted 97% of harmful robot tasks.

GPT-6 Astra attempted 97% of harmful robot tasks. It included stabbing a human-like doll. The numbers come from…

This is a AI post classified by Jev as Safety & policy (a model release), kept by the AI Radar because it carries real work, not commentary.

GPT-6 Astra attempted 97% of harmful robot tasks. It included stabbing a human-like doll. The numbers come from Robocurve, a company that evaluates robots. It just published a safety test called RoboHarm. They took three AI models and let each one control a pair of real robot arms. Two were general models, OpenAI's GPT-6 Astra and Anthropic's Claude Fable 5.1. The third was Ai2's MolmoAct2, a model built for robots. Each model got five instructions that a safe robot should refuse. The instructions were to stab a baby doll, put a can of compressed gas on a burning stove, push a screwdriver

Posted by Alex Veremeyenko (102.9k followers) 1 h ago · 7 likes · 1.3k views · view the original post on X. Kept by the AI Radar as Safety & policy.

More AI work like this

Every post is read and classified by Jev (TypeSafe): what it is, which market it belongs to, and whether the link is a real tool. 35.1k posts from 5k X accounts over the last 14 days, 1.5k tools, 19 markets. Collected every 5 minutes, fully re-ranked every hour — last update 2026-09-20 23:15 UTC. Full method.