Anthropic confirms Bay Area wet lab where its AI runs real biology experiments

Anthropic confirms a Bay Area wet lab

Anthropic has confirmed that it operates a wet biology laboratory in the Bay Area, where it can use its AI models to run physical experiments rather than simulations. The company acknowledged the facility to TechCrunch, making it one of the most concrete examples yet of a frontier AI lab moving beyond computational prediction into hands-on experimental science.

Eric Kauderer-Abrams, Anthropic's head of life sciences, told Reuters that "to do biology, the final test is still, and will be for a while, in real lab work," and that Anthropic is absolutely doing that today. According to TechCrunch, the facility works much like an ordinary biotech lab: Anthropic runs some research in-house while also collaborating with external partners.

The reveal is not entirely out of the blue. As TechCrunch notes, Anthropic acquired Coefficient Bio, a stealth AI biotech startup, in April, so the existence of physical lab capability is unsurprising in hindsight.

Fundamental biology, not drug discovery

Anthropic declined to detail what the lab is working on, but said its main focus is fundamental biology rather than drug discovery. TechCrunch reads that emphasis as deliberate positioning. Anthropic counts major pharmaceutical companies among its customers and partners, including a newly announced joint drug-discovery agreement with Novo Nordisk, and it has already faced backlash for launching products perceived to compete with its own customers.

Alongside the lab news, Anthropic launched a Life Sciences Verification Program this week, which gives vetted biology researchers access to its most powerful models. It has also recently published reports on accelerating protein design and on biomolecular modeling as part of its work supporting drug discovery.

Why an AI company needs benches and pipettes

CEO Dario Amodei said last week that he believes AI "could cure most major diseases in the next 5-10 years." A claim like that requires more than models that generate plausible hypotheses: the hypotheses have to be tested against physical reality. A wet lab gives Anthropic a closed feedback loop in which its models can propose experiments, observe real results and revise their understanding, much as iterative feedback has driven progress in AI-assisted coding, but applied to cells and molecules instead of software.

A timing problem

The announcement lands during an unusually fraught stretch for Anthropic's public safety messaging. TechCrunch recalls the recent resignation of researcher Jacob Coxon, who warned that "the people building AI earnestly believe that it could kill us all by the end of the decade," and notes that Anthropic's own alignment lead has put the odds of AI exterminating humanity within the next decade at greater than 10 percent. Last weekend, Amodei published a post urging the industry to slow down and adopt self-regulation, and he has repeatedly named bioterrorism as one of AI's biggest risks.

The contrast has not gone unnoticed in the tech industry. Investor and AI coding startup founder Chamath Palihapitiya joked on X that the company behind such hits as "We're All Going To Die" and "Regulate Me Now" is now building a wet lab in San Francisco, concluding: "I do not recommend this."

Why it matters

Most AI-for-science efforts to date have stopped at computation, predicting protein structures or proposing candidate molecules that human researchers must then validate. A frontier AI lab operating its own wet lab, with its models inside the experimental loop, is a structural step toward genuinely AI-driven discovery and toward the disease-curing ambition Amodei keeps invoking. It also blurs the line between AI vendor and research organization, which has commercial consequences given Anthropic's pharmaceutical customer base. And it puts the company's own safety rhetoric to a practical test: a firm whose CEO calls bioterrorism a top AI risk and asks the industry to slow down is now physically running biology experiments with its models. How Anthropic governs, staffs and reports on that work will be watched closely by both its partners and its critics.