Anthropic is setting up a biology lab where Claude guides robots through drug experiments

A wet lab for AI drug development

Anthropic is setting up its own biology laboratory, moving its AI-driven drug development work beyond computer simulations. The lab is located in the San Francisco area and will let the company run physical experiments, Reuters reports. The plan is for Claude to guide robots through those experiments with minimal human involvement.

Two tools already shipped

To make that workflow possible, the company has already released two pieces of infrastructure: Claude Science, an AI workspace built specifically for researchers, and the Model Hardware Standard, an effort to control laboratory equipment in the way its Model Context Protocol connects AI systems to software. Human oversight is still required for safety, Anthropic says.

Focus on so-called undruggable diseases

Eric Kauderer-Abrams, who runs Anthropic's life sciences division, sees the biggest opportunity in diseases long considered "undruggable" because no treatment exists. In his view, AI can speed up the design of complex antibodies that hit several targets at once — work that is difficult to carry out by hand.

Earlier moves this year

The lab builds on steps the company took earlier this year. In April it acquired Coefficient Bio, a young AI pharma startup with fewer than ten employees, in a deal worth about $400 million in shares, and added Vas Narasimhan, the chief executive of Novartis, to its board. According to the report, Anthropic will not run its own clinical trials for now, an approach that avoids competing directly with pharmaceutical companies.