AI Agent Controls Brain Injection in Live Mouse After Claude Opus Refuses

An AI agent at Exin Therapeutics controlled robotic equipment to inject a gene therapy into the brain of a live mouse, according to WIRED. A veterinary surgeon first performed the craniotomy while the agent observed.

The agent selected a treatment from 19 options, chose the brain region and injection coordinates, and controlled the robotic arms, needle, volume, and injection speed. Anthropic's Claude Opus 5.5 completed the planning but refused to guide the needle into the brain, so the team switched to Claude Sonnet 5.

The researchers required human approval of the injection plan, installed an emergency stop button, and blocked the agent from changing the equipment code. When a glitch disrupted the coordinate system, the agent recognized the error but would not change the code, so the team used a separate agent running Opus 5 to fix it without telling it a live surgery was under way. The delay extended the procedure to slightly more than an hour, compared with about 45 minutes normally. The mouse is recovering, and the team is waiting to assess whether the therapy reduces its seizures. Anthropic did not respond to a request for comment, WIRED reports.