Anthropic, the artificial intelligence company led by CEO Dario Amodei, has established a physical biology laboratory in the San Francisco Bay Area as it expands its work in life sciences. The move marks a shift beyond computer-based research and reflects the company’s broader effort to apply advanced AI systems to biological research and medical science.
According to a Reuters report published on September 18, 2026, Anthropic has quietly established a wet laboratory where physical biological experiments can be conducted. Two people familiar with the company’s activities confirmed the existence of the facility. Anthropic’s head of life sciences, Eric Kauderer-Abrams, also confirmed that the company is conducting real laboratory work.
However, Anthropic has clarified an important point about the facility. A company spokesperson said the laboratory is not specifically intended for drug discovery, although the company is developing a wider programme focused on applying artificial intelligence to life sciences and medical research. This distinction is important because Anthropic’s work currently covers a broader range of biological research rather than functioning as a conventional pharmaceutical company.
Anthropic has said that it sees significant potential for AI to accelerate scientific research and help improve the detection and treatment of diseases. Its stated areas of interest include rare diseases and medical conditions that may receive less attention from traditional pharmaceutical companies because of the financial and scientific challenges involved.
The company’s move into physical laboratory work represents an expansion of its previous approach, which largely relied on AI models performing research and analysis through computers. Anthropic’s life sciences team now combines computational work with physical experiments conducted either in its own facilities or through external partners.
Kauderer-Abrams said real laboratory testing remains an important part of biological research. The company therefore believes that AI-generated scientific insights need to be evaluated through physical experiments rather than relying entirely on computer simulations.
One of Anthropic’s longer-term goals is to explore whether its Claude AI systems can assist with the automation of laboratory processes. According to Reuters, the company is interested in using Claude to direct robotic systems and laboratory equipment to carry out scientific experiments with limited human intervention. The company has stressed that human involvement and oversight remain important for safety.
This effort comes as AI companies increasingly explore applications beyond conventional software. Artificial intelligence is already being used in areas such as protein analysis, molecular research, medical data analysis and drug-development research. Anthropic’s expansion suggests that AI companies may increasingly become involved in the physical processes required to test scientific hypotheses.
Anthropic has also been developing tools and programmes aimed at supporting life sciences researchers. On September 17, 2026, the company announced its Life Sciences Verification Program, which provides verified life science organisations with access to selected Anthropic models under specialised safeguards. The programme is designed to support areas including drug discovery, research biology, clinical development and manufacturing.
The company said organisations participating in the programme must go through a verification process involving research credentials, security standards and ethical research oversight. Anthropic said the programme is intended to provide researchers with greater access to advanced AI capabilities while maintaining safeguards around biological applications.
Anthropic has also acknowledged that advanced AI models can create dual-use risks in biology. In August, the company said some biological capabilities could be useful for legitimate medical research while potentially creating risks if misused. As a result, it has been developing additional safeguards and trusted access systems for advanced biology-related applications.
The company’s investment in life sciences is therefore taking place alongside its efforts to address AI safety concerns. Anthropic has previously warned that increasingly capable AI systems could present serious risks if their capabilities are misused. Its biology programme is being developed within that broader framework, with the company saying that human oversight is necessary when AI systems interact with real-world laboratory environments.
Anthropic has also been building partnerships and acquiring technology to strengthen its position in life sciences. Reuters reported that the company acquired biotechnology startup Coefficient Bio, a move intended to support the development of tools for biological research and drug development. The company has also expanded its relationships with pharmaceutical and biotechnology organisations.
At the same time, Anthropic says it is not currently attempting to compete directly with pharmaceutical companies in bringing finished medicines to patients. Eric Kauderer-Abrams told Reuters that the company is not conducting clinical trials at this stage. Instead, its focus is on earlier stages of research where AI could potentially help scientists identify new approaches to difficult biological problems.
The potential benefits of this approach are significant but remain uncertain. Developing a successful medicine typically requires years of research, laboratory testing and clinical trials. Identifying a promising biological target or molecule does not guarantee that it will become a safe and effective treatment. Most potential medicines do not ultimately reach the market because of challenges involving safety, effectiveness, manufacturing or other factors.
Anthropic’s work could therefore be viewed as an early-stage attempt to determine how advanced AI can contribute to biological research rather than evidence that AI has already solved major medical problems.
The company is particularly interested in conditions that researchers have historically considered difficult to treat. Its scientists are examining whether AI could help accelerate areas such as complex antibody development and other forms of biological research. The company has not publicly disclosed the precise diseases or research programmes being pursued in its laboratory.
Anthropic’s expansion into physical biology also raises questions about how AI companies should manage sensitive scientific information. Pharmaceutical companies and biotechnology firms often work with confidential research data, proprietary drug candidates and commercially sensitive discoveries. AI companies working with these organisations therefore need to maintain strong controls over data access and model training.
Anthropic has said that information handled through its life sciences programmes is compartmentalised and cannot be used for model training or accessed by members of its broader life sciences research teams under the stated programme rules.
The development of a physical laboratory demonstrates how the AI industry is increasingly moving from digital environments into real-world scientific settings. Instead of using AI only to analyse information or generate predictions, companies are exploring systems that can interact with laboratory instruments, robots and biological experiments.
For Anthropic, the objective is to determine whether advanced AI can accelerate scientific progress while maintaining appropriate safeguards. The company’s Bay Area laboratory is an early step in that process. Its success will depend not only on the capabilities of AI models but also on the quality of biological experiments, scientific validation, safety controls and eventual clinical evidence.
At present, Anthropic’s biology initiative remains in an early stage. The company has confirmed its wet-lab activities, expanded its life sciences programmes and indicated that it wants AI to contribute to medical research. However, there is no indication that Anthropic has developed a commercially available medicine through this programme or that its research has reached human clinical trials.
The company’s latest move nevertheless illustrates the growing intersection between artificial intelligence and biotechnology. As Anthropic continues developing Claude and its life sciences tools, its laboratory work could provide further insight into how AI can be combined with physical experimentation to accelerate research into diseases and potential treatments.