Anthropic's research page is dated Oct 27, 2026. The life sciences team writes that, over the past several months, it has worked with six organizations developing new medicines: BigHat Biosciences, Bristol Myers Squibb, Divergence Labs, Enveda, Generate Biomedicines, and The Column Group. The end of the page says the research is still early and, so far, mostly preclinical, meaning most of it is not yet at the stage of testing the drugs in humans. The figures below are lab results, paperwork results, or the company's own expectation, as the page reports them. They are not an independent clinical-trial registry.
[1]
The Generate Biomedicines section says six scientists used Claude Code to propose new designs. One design was about 70 percent shorter than the standard version. In lab tests, immune cells made roughly 10 times as many receptors, killed tumor cells more efficiently, and stayed active longer. The page presents this as a laboratory test. It does not present it as something already given to patients.
[1]The Divergence Labs section says roughly 35,000 antibody designs were generated in about a month and then narrowed for testing. The early tests found three designs that could bind IgE and 20 that could dislodge it, and three of those 20 appeared to dislodge it more than the standard medication. The page is reporting those test counts. How a molecule would be assembled is not part of this account.
[1]The BigHat section says the lab tested about 1,010 variants in about 10 weeks. Compared with the original leads, the best versions bound more than 10 times more tightly, the cancer-targeting lead stayed in solution 18 degrees Celsius higher, and the lead aimed at keeping a drug in the body longer clumped more than 100 times more slowly. Those are the laboratory comparisons as the page states them. The steps used in the comparison sit in a technical report, and they are not repeated here.
[1]The Enveda section is paperwork, not manufacturing. The page says that across tests on 1,061 pages of documents, Claude caught all 11 issues human reviewers had marked and flagged seven they had missed. Human review time fell by 89 percent, from over 25 hours to less than three. Two regulatory drafts are separate. One took four hours, against about three months by hand. Enveda's ordinary human review found one minor error in the other. The page does not say those documents have been approved.
[1]The Column Group is setting up Pleiades. The page says the first skill for judging whether a target is worth pursuing has been evaluated on 45 target-disease pairs. That is not a medicine. Michael Pourdehnad, who leads early clinical development at Bristol Myers Squibb, says on the page that one prompt produced an initial analysis that would otherwise take days. The page says the two sides plan to expand the work in 2027. It does not describe how a trial plan is written.
[1]The page also says Anthropic acquired Mithril Medicine. Founder Cody Tranbarger took a possible treatment for one form of muscular dystrophy from an idea to results in mice in less than 18 months. The page then says that if the candidates continue to hold up, the company expects a drug could be ready to start clinical trials in humans as soon as late 2028. That is the company's expectation inside a conditional sentence. Separately, the page says the company's own programs include rare childhood forms of muscular dystrophy, and that Claude handles much of the scientific and operational work there. That description is the company's, and it is separate from the six partner results above. The same section says again that most of this work is not in human testing. The page does not name an independent lab that has repeated the results, and it does not say that any program has regulatory approval.
[1]