Anthropic Enters the Drug Discovery Race, Launches Claude Science Platform

Anthropic, the company known primarily for its Claude models, has announced its entry into biotechnology. The company has launched Claude Science, a tool for researchers working in biomedical research, and at the same time announced that it will begin running its own drug discovery programs. It is Anthropic's first such direct step into actual laboratory research, rather than simply supplying models to other companies.

What Is Claude Science

Claude Science is an integrated work environment meant to replace the scattered set of tools research teams use today. The platform allows all stages of scientific work to be carried out in one place, from literature review to preparing a publication-ready manuscript, and includes more than 60 built-in functions covering genomics, proteomics, single-cell analysis and cheminformatics. It integrates with databases including UniProt, PDB and Ensembl, and for working with microbial sequences it draws on the EDEN dataset from Basecamp Research, described as the largest database of its kind in the world.

Reproducibility of results is a key element. Every generated chart or protein structure comes bundled with the code that created it, along with the full history of the method and the conversation context. Computations can be run on a laptop, a lab's computing cluster, or on-demand rented GPUs, and sensitive research data does not need to leave the systems where it already resides. The tool is currently in beta for users on the Pro, Max, Team and Enterprise plans, on macOS and Linux.

Anthropic's Own Drug Program

Alongside the Claude Science launch, Anthropic announced that it will itself begin running drug discovery programs, initially focused on neglected diseases, meaning conditions traditional pharmaceutical companies avoid because they see no attractive return on investment. Eric Kauderer-Abrams, Anthropic's head of life sciences, explains the move as part of a search for what the company should be doing beyond training models and building products. The decision fits with Anthropic's status as a public benefit corporation, which formally allows the company to place patient welfare above profit.

For Anthropic itself, this is also a way to test its own models in practice. The company speaks of a "tight feedback loop" between AI development and real scientific work, meaning a situation in which lab experience directly shapes further development of Claude. Outside researchers can also benefit from the Claude Science launch: up to 50 research projects will receive credits worth up to $30,000, applications are accepted through July 15, 2026, and results are expected to be announced by the end of the month.

The Third Player in the Race

Anthropic's entry into drug discovery means that the three biggest players in the large language model market, Anthropic, OpenAI and Google, are now running parallel programs aimed at the life sciences and pharmaceutical sector. Each company is betting on a somewhat different model of collaboration with researchers, but all share the belief that language models can shorten the time needed to analyze biological data from years to weeks. The launch coincided with an earlier decision by Washington to lift some export restrictions on Anthropic's models, making it easier for international research teams to access the latest versions of Claude.

Scientists testing the platform in beta praise the time savings above all. Jérôme Lecoq of the Allen Institute notes that Claude Science made it possible to carry out the entire process of preparing a scientific review comprehensively, in far less time than the roughly two years such work typically takes a team. Stephen Francis of UCSF, meanwhile, describes a dramatic speedup in analyses, cutting the working time down to about one-tenth of what it used to take.

What This Means for Science

For Polish research institutions and biotech companies, the launch of Claude Science signals that AI tools for analyzing biological data are no longer the domain of only the largest pharmaceutical labs. The grant program, offering credits of up to $30,000, is open internationally, so teams from Polish universities and institutes can also apply, as long as they meet the submission deadline.

At the same time, it is worth remembering that claims about accelerating drug discovery are heard regularly from AI companies, and the real test only comes once one of a model's proposed drug candidates actually reaches clinical trials. Anthropic has not yet disclosed exactly which neglected diseases its own program will target, or whether it intends to commercialize the results, so a genuine assessment of this initiative's impact will only be possible months or years from now.

Sources: CNBC (cnbc.com), STAT News (statnews.com), pharmaphorum (pharmaphorum.com), Anthropic Newsroom (anthropic.com).