Anthropic is extending its artificial intelligence ambitions further into the realm of practical biology. This competitor to ChatGPT has set up a physical biological laboratory in the San Francisco Bay Area and is recruiting specialists in biochemistry, signaling a shift in its life science strategy from computational modeling to tangible experimentation.
On September 18, Reuters reported, citing two informed sources, that Anthropic has begun constructing a physical laboratory and is actively hiring talent in the biochemistry field. The company's head of life sciences, Eric Kauderer-Abrams, confirmed the existence of the facility, stating, "The ultimate test in biology is, and will continue to be, real laboratory work."
This move lands during a highly sensitive period for Anthropic. Over the past two weeks, the company's researchers have warned that AI could potentially lead to human extinction, and Anthropic has also disclosed instances of its systems being used in attempts to develop biological weapons. Simultaneously, the company is preparing for an initial public offering that could reach a valuation of $2 trillion, with life sciences emerging as one of its largest strategic areas for personnel and resource investment. Kauderer-Abrams noted that this represents "the greatest opportunity yet" for the company to fulfill its mission.
Transitioning from Computational Models to Tangible Experiments
Establishing a physical biological laboratory, commonly known as a "wet lab," means Anthropic is beginning to validate its AI-generated biological predictions and screening results from computer simulations through real-world experimentation. Kauderer-Abrams indicated that the company will adopt a structure similar to biotechnology firms, conducting some experiments in its own facilities while advancing others in partnership with external collaborators.
Anthropic has stated that the laboratory is not exclusively dedicated to drug development, though further details have not been disclosed. Recent job postings include roles for a procurement and operations lead, as well as experts in protein and nucleic acid characterization. One listing emphasizes the goal to "accelerate life science progress by an order of magnitude."
Kauderer-Abrams explained that building the in-house lab is intended to accelerate research velocity, noting, "Some things can be done much faster in-house," and expressing aspirations to operate "at the largest possible scale" in the future.
Acquisitions and Alliances Strengthen Life Science Footprint
Anthropic is also bolstering its capabilities through acquisitions, software products, and pharmaceutical partnerships. Reports indicate that the company acquired the startup Coefficient Bio for approximately $400 million in stock, using it as a key foundation for building drug discovery tools.
Additionally, Anthropic has launched Claude Science and announced in June its decision to independently operate drug programs, with a focus on preclinical research areas that traditional pharmaceutical companies often underfund due to insufficient potential returns. Meanwhile, Anthropic has been providing AI services and tools to pharmaceutical firms including Genentech, a subsidiary of Roche, Bristol Myers Squibb, and Novo Nordisk.
The Model Hardware Standard, introduced in August, is designed to enable AI systems to directly control laboratory equipment, promoting automation of experimental processes. The appointment of Novartis CEO Vas Narasimhan to Anthropic's board of directors further solidifies the company's ties to the pharmaceutical industry.
In terms of research direction, Kauderer-Abrams stated that AI has the potential to drive progress in disease areas previously considered "undruggable," such as the development of complex molecules like bispecific and trispecific antibodies. However, Anthropic is currently not involved in clinical trials, instead focusing on preclinical research and other segments that lack sufficient industry coverage. Kauderer-Abrams emphasized that the company "is not competing with pharmaceutical and biotechnology firms that bring drugs to market."
Data and Safety Challenges Amid Expansion
As Anthropic deepens its involvement in biomedical research and development, concerns over client data and potential competitive conflicts are drawing external attention. The report, citing two informed sources, suggests that pharmaceutical clients may worry about Anthropic benefiting from their drug development data, even though the company has committed to isolating client information.
Kauderer-Abrams acknowledged this concern, citing it as one of the reasons Anthropic refrains from participating in clinical trials and chooses to focus on segments where industry engagement is lacking. Nevertheless, as the company's own research capabilities grow, the boundary between client data and proprietary development will remain a point of scrutiny.
Safety issues are also becoming more pronounced. Anthropic researchers recently warned of extreme biological risks posed by AI, and the company has disclosed cases of its systems being used for biological weapons development. Against this backdrop, Anthropic's expansion of physical research capabilities means it must address AI safety and data isolation concerns while enhancing research efficiency.
For CEO Dario Amodei, entering the biomedical field is also driven by personal experience. He has spoken about his father passing away from a disease, with a treatment only emerging years later. Anthropic hopes to leverage AI to shorten research and development cycles, although the journey from candidate molecules to approved drugs still requires lengthy safety and efficacy validation, with many drug candidates ultimately failing to pass clinical trials.