A team led by Kexin Huang, a professor of engineering at Stanford University in the United States, has developed Biomni, an artificial intelligence (AI) agent that automates the entire process of biomedical science research from literature search to experiment design, and published the results in the international journal Science on the 9th (local time). Biomni analyzes the literature to identify databases, software, and other work resources, and a large language model (LLM) breaks down the user's request into multiple sub-tasks and draws up an execution plan. In five case studies—including analysis of wearable sensor data, design of experimental protocols, and optimization of protein stability—it demonstrated the ability to perform multi-step workflows. The research team expects Biomni to allow researchers to hand over repetitive tasks to AI. Lim Soo-jong, head of the Language Intelligence Research Laboratory at the Electronics and Telecommunications Research Institute (ETRI), said the core is that an LLM carries out the entire research process as a single workflow.