AI agent performs gene-therapy injection in mouse brain

WIRED reports that a biotech startup let an AI agent choose a gene therapy and successfully inject the treatment into a live mouse’s brain. The experiment offers a striking glimpse of how autonomous agents could participate in scientific research, from selecting candidates to carrying out delicate procedures. It also raises questions about oversight, reliability and accountability when AI systems are given control over biological experiments. The significance lies in the combination of decision-making and physical action. An AI agent that can evaluate options and then execute an experimental step moves beyond analysis and into the lab itself. In this case, the agent reportedly selected a gene therapy and completed an injection into a mouse brain. Such work could accelerate research by automating repetitive or complex tasks, and it may help scientists explore more candidates in less time. The startup’s work suggests that agentic AI could become a useful partner in drug discovery and preclinical testing. But the experiment also illustrates the risks. Biological procedures require precision, safety checks and ethical review. If an AI agent is involved in choosing treatments or performing injections, researchers need to understand its failure modes and ensure human accountability. Questions include how the agent’s choices were validated, what safeguards prevented harm and who would be responsible if something went wrong. The mouse study is not a sign that AI can replace scientists, but it does show that the boundary between computational assistance and autonomous lab work is blurring. As agentic tools become more capable, oversight frameworks will need to keep pace with the experiments they enable.