Anthropic, one of the leading artificial-intelligence companies and the maker of the Claude chatbot, is moving into an unexpected line of work: trying to discover drugs. The company said it is beginning an internal drug-discovery program and, at the same time, releasing a product called Claude Science aimed at researchers in the pharmaceutical industry, STAT News reported.
What Anthropic is doing
The announcement has two parts. The first is a tool: Claude Science is a version of Anthropic's AI tuned for scientific research, designed to help with stages of the drug-development process such as identifying promising biological targets, refining candidate compounds, and making sense of complex biology, according to CNBC. It is pitched at pharmaceutical companies and academic scientists as an assistant for the painstaking work of research.
The second, more striking part is that Anthropic intends to try using these tools itself, by running its own drug-discovery effort in-house. The company's head of life sciences, Eric Kauderer-Abrams, said the rationale is partly about credibility: to build the right models and products for the industry, Anthropic believes it needs to "live it" alongside the scientists it hopes to serve — in other words, to feel the difficulties of real research rather than just supply software from the outside.
A focus on 'neglected' diseases
Anthropic has said its in-house effort will concentrate on so-called neglected diseases — conditions that traditional biopharmaceutical companies tend to avoid because they are not commercially attractive, often because they mainly affect poorer populations or too few patients to promise a large return. Directing AI-assisted discovery at those targets is a notable choice, and one the company can present as socially useful as well as strategically motivated.
Whether Anthropic aims to take any resulting drug candidates all the way to market, however, is not yet clear. The near-term emphasis, executives suggested, is on gaining hands-on experience and improving the tools, rather than on becoming a conventional drugmaker overnight.
Why it matters, and the caveats
The move fits a broader trend of AI companies pushing into science, betting that their models can accelerate discovery in fields like biology and chemistry. The appeal is enormous: drug development is famously slow, expensive and prone to failure, and even modest improvements in speed or hit rate could be hugely valuable, both commercially and for patients.
But the caveats are equally real. Turning a promising computational lead into an approved medicine takes many years of laboratory work, animal studies and human clinical trials, most of which still fail. AI can help sift possibilities and generate hypotheses, but it cannot shortcut the biological reality of testing whether a compound is safe and effective in people. History is littered with technologies that promised to revolutionize drug discovery and delivered more slowly than their boosters hoped.
There is also a question of expectations. AI firms have strong incentives to showcase ambitious, science-changing projects, and announcements can outrun results. The meaningful test of Anthropic's effort will not be the launch but what, if anything, comes out of it over the years ahead.
The bigger picture
For Anthropic, the initiative is a statement of ambition — a claim that its AI is not just a chatbot but a tool capable of contributing to hard scientific problems. For the pharmaceutical industry, it is another sign that the world's most powerful AI developers intend to be participants in research, not merely vendors to it.
If the approach works even partially, it could help direct attention and resources toward diseases that have long been under-served. If it does not, it will join a long line of high-hopes ventures at the intersection of computing and medicine. Either way, it marks a notable step in the spread of artificial intelligence from the screen into the laboratory — and, its backers hope, eventually into the medicine cabinet.
