AI Agent Selects Gene Therapy and Injects It Into a Live Mouse's Brain
One-line summary: A biotech startup let an AI agent pick a gene therapy, then had it inject the treatment into a live mouse's brain, and the injection succeeded.
Key points
- The agent itself chose which gene therapy to use, rather than a human researcher.
- It carried the work through to injecting the chosen treatment into the brain of a living mouse, which was reported as successful.
- The test subject was a live animal, so this goes beyond research AI that stays with literature search or data analysis and moves into physical experiments.
- Wired framed the result as a startling glimpse of what future research labs could look like.
Why it matters
If AI moves from proposing hypotheses to selecting treatments and carrying out experiments, research speed and staffing models could change significantly. At the same time, handing in-vivo work to an agent raises questions about safety validation, accountability, and oversight that need to be addressed alongside the technology.