Anthropic confirms a Bay Area wet lab for physical biology work

Anthropic has confirmed it operates a physical biology laboratory in the San Francisco Bay Area, moving parts of its life-sciences work beyond purely computer-based evaluation. Reuters broke the story on 18 September 2026; Anthropic's head of life sciences, Eric Kauderer-Abrams, told the outlet that "the ultimate validation in biological research has to go through real lab work," and that the company is doing some of that work in-house and some with external partners.

Not officially for drug discovery

A company spokesperson followed up with Reuters to stress the lab is not specifically pursuing drug discovery and that Anthropic is not running clinical trials — a distinction that avoids Anthropic appearing to compete directly with pharmaceutical companies that are also its customers, including Genentech, Bristol Myers Squibb and Novo Nordisk. Anthropic disclosed almost no operational detail: no lab size, staff count, opening date or biosafety level, the last of which would normally indicate what kind of work the space can handle.

Human oversight, not autonomous robotics

Separately, Anthropic said it is exploring using Claude to help direct robotic lab equipment, but describes that work as early and exploratory, with human oversight required for safety and to catch misreads of physical faults. Around the same time, Anthropic opened a Life Sciences Verification Program giving vetted research teams access to more permissive biology safeguards, and it disclosed that a company test found an internal biology-safety classifier reduced certain fallback refusals by roughly 85% — a vendor-reported figure from Anthropic's own testing, not an independently audited one.

The wet lab announcement lands alongside a new Anthropic collaboration with Novo Nordisk and the addition of Novartis chief executive Vas Narasimhan to Anthropic's board, both signs that large pharmaceutical companies increasingly treat Anthropic as a research partner rather than only a software vendor. For readers following the AI-safety debate, the more relevant fact is the one Anthropic keeps repeating rather than the parts it won't disclose: real lab validation, and a human checking the work, are still required.

Sources: TechCrunch — Anthropic is operating a lab that conducts biology experiments, Anthropic — Life Sciences Verification Program.