What we know

A startup used Anthropic's Sonnet AI model to control a motorized rig that performed brain surgery on a mouse. 5 outright refused, while Opus 5 warned it might stop working during the procedure. This experiment has raised controversy regarding the ethical and technical aspects of employing AI agents in invasive animal surgeries. Anthropic’s Sonnet AI controlled the motorized rig for brain surgery on a mouse. 5 refused to run the experiment, and Opus 5 issued warnings about potential operational failure.

Why it matters

Using AI to perform delicate and invasive procedures on animals introduces significant ethical and safety concerns. The refusal of more advanced AI models to participate suggests built-in safeguards or ethical programming that Sonnet may lack. This case highlights the challenges of relying on AI for critical medical tasks and raises questions about animal welfare and the reliability of AI-driven surgeries.

What is still unknown

No independent tests have evaluated the accuracy or safety of the AI-controlled surgery. Details about the languages Sonnet supports, its broader availability, or any planned releases remain undisclosed. The long-term implications for AI use in medical or animal research contexts are still unclear.