Anthropic Acquires One-Person Pharma Startup and Partners with Six Major Institutions for Full-Scale AI Drug Discovery Push
nashnova research
Anthropic acquired Mithril Medicine — a rare-disease biotech run by a single founder — and forged research partnerships with six institutions, deploying its AI model Claude across the full drug-development pipeline from molecule design to clinical planning, with human trials targeted as early as late 2028.
Why acquire a company run by one person?
Mithril Medicine was operated solely by founder Cody Tranbarger. It focused on two rare childhood muscular dystrophies. Tranbarger delegated contract management, lab coordination, and much of daily operations to Claude.
The result: in under 18 months, one candidate therapy moved from concept to mouse studies with early positive results. This means → one person plus one AI matched the development pace of a small team.
Tranbarger has joined Anthropic to continue the work and launch new projects. Anthropic says that if the candidate drug keeps performing, human clinical trials could begin as early as late 2028. Deal terms were not disclosed.
What are the six partners doing?
Generate Biomedicines is trying to extend CAR-T cell therapy — a technique that reprograms a patient's own immune cells to attack cancer — from blood cancers to solid tumors. Six scientists used Claude Code to propose new designs; one was ~70% shorter than the standard version, and lab tests showed immune cells producing roughly 10× more receptors, with higher tumor-killing efficiency and longer activity.
Divergence Labs is developing a new drug for food allergies. Claude near-autonomously generated roughly 35,000 antibody designs in about a month; early screening identified 3 candidates that bind IgE — the molecule that triggers allergic reactions — and 3 outperformed existing standard drugs.
BigHat Biosciences combines AI with automated labs to optimize drug-candidate molecules. In roughly 10 weeks the team tested about 1,010 molecular variants; the best version showed binding affinity up over 10×, thermal stability of a cancer-targeting candidate rose by 18°C, and aggregation rate of another candidate dropped over 100×.
How much faster can document review and regulatory filing get?
Manufacturing chemists at Enveda previously spent up to 75% of their time on document work. After connecting their document pipeline to Claude, a test on 1,061 pages showed Claude caught all 11 issues human reviewers found — and flagged 7 more they missed.
In plain terms = the AI didn't just match human review — it was more thorough. Total manual review time fell from over 25 hours to under 3 hours, an 89% reduction.
Claude also drafted two key regulatory documents; one took 4 hours — the same work typically requires about 3 months by hand. This means → regulatory paperwork, traditionally the slowest bottleneck, is being compressed from months to days.
How are big pharma and venture capital involved?
Venture firm The Column Group (TCG) partnered with Anthropic to create a new company, Pleiades, designed to run more and more complex drug programs with small teams. Pleiades has built its first Claude skill module and completed assessments across 45 target-disease pairings.
Bristol Myers Squibb is the newest partner. Senior VP Michael Pourdehnad said a single prompt produced a complete preliminary analysis for evaluating development options — work that would normally take days. The two sides plan to expand the collaboration in 2027.
This reflects a shift: AI-driven drug discovery has moved beyond startup experiments to the stage where major pharmaceutical companies are willing to formally partner.
What else is Anthropic building around this?
Anthropic launched Claude Science, an AI workbench for researchers, alongside a Life Sciences Validation Program that gives trusted institutions access to its most powerful models under relaxed safety constraints.
The company is also openly recruiting visiting researchers for scientific teams and offering Claude usage credits for life-science researchers. Put simply = Anthropic isn't just making drugs itself — it aims to turn Claude into foundational infrastructure for the entire industry.
Anthropic stresses that all research remains at an early stage; most has not yet reached human trials. Whether human clinical trials can launch on schedule by late 2028 will be the critical test of this biotech strategy's viability.
市场有风险，内容仅供研究参考，不构成投资建议。