Anthropic has published early results from six drug developers using Claude across the pipeline, from designing molecules to drafting regulatory documents. All the figures come from Anthropic and its partners, and most of the work is preclinical.
At Generate Biomedicines, Claude-proposed designs produced a component of a CAR T-cell therapy about 70% shorter than the standard version; in lab tests the cells made roughly ten times as many receptors and killed tumour cells more efficiently. At BigHat Biosciences, the lab made and tested about 1,010 antibody variants, and the best bound more than ten times more tightly than the originals. At Enveda, Claude caught all 11 issues human reviewers had found in 1,061 pages of manufacturing documents and flagged seven more, cutting human review time by 89%. Bristol Myers Squibb is using it in clinical trial planning.
Anthropic also says it has acquired Mithril Medicine, a one-person company working on treatments for two rare forms of muscular dystrophy, whose work went from idea to promising results in mice in under 18 months.
Why it matters: drug development is slow and expensive, and these are named companies with measured results rather than promises. But no drug here has been tested in people yet, and the results are reported by the companies rather than peer-reviewed studies.
Source: Anthropic