Summary
Anthropic's life-sciences team published a research blog on October 27, 2026, detailing early progress on Claude-accelerated drug discovery: partnerships with BigHat, BMS, Divergence, Enveda, Generate and The Column Group, plus in-house discovery programs, with human clinical trials possible as early as late 2028.
Overall Progress
What drug-discovery milestones did Anthropic disclose?
Per Anthropic's official research blog (October 27, 2026), its life-sciences team deepened collaboration over recent months with six institutions: BigHat Biosciences, Bristol Myers Squibb, Divergence Labs, Enveda, Generate Biomedicines and The Column Group, plus in-house drug discovery programs. Supporting moves include launching the Claude Science research workbench, providing Claude credits to life scientists, and setting up the Life Sciences Verification Program.
Which part of drug R&D does each of the six partners cover?
Per Anthropic's blog, the six partners are BigHat Biosciences, Bristol Myers Squibb, Divergence Labs, Enveda, Generate Biomedicines and The Column Group. Generate works on CAR-T redesign, Divergence on food-allergy antibodies, BigHat on antibody optimization with automated labs, and Enveda on quality-document review; the Bristol Myers Squibb collaboration expands in 2027, while The Column Group's specific role is not detailed in the blog.
Concrete Results
What data did Generate's solid-tumor CAR-T redesign produce?
Per Anthropic's blog, six scientists at Generate Biomedicines used Claude Code to design CAR-T genetic instruction segments: one design was about 70% shorter than the standard version, lab tests showed immune-cell receptor production up about 10x, and tumor-killing efficiency was higher.
How far has Divergence's food-allergy program progressed?
Per Anthropic's blog, in Divergence Labs' food-allergy program, Claude autonomously generated about 35,000 antibody designs in roughly one month. Early testing yielded 3 IgE-binding designs and 20 IgE-stripping designs, of which 3 stripped IgE better than the standard drug.
How much did experimental efficiency improve at BigHat and Enveda?
Per Anthropic's blog, BigHat and Claude, linked with an automated lab, manufactured and tested about 1,010 variants over about 10 weeks: the best version showed over 10x stronger binding, tumor-targeting antibodies tolerating 18°C higher temperatures, and long-acting candidates aggregating over 100x more slowly. Per the same blog, in a 1,061-page quality-document test at Enveda, Claude caught all 11 issues found by human reviewers plus 7 more; human review time dropped from over 25 hours to under 3 hours (an 89% reduction), and a regulatory draft took only 4 hours versus about 3 months manually.
In-House Pipeline and Timeline
What is Mithril Medicine, the company Anthropic acquired, and what is the clinical timeline?
Per Anthropic's blog, Anthropic acquired the one-person company Mithril Medicine, focused on rare childhood muscular dystrophy. Its candidate drug went from idea to positive mouse results in under 18 months, and human clinical trials could start as early as late 2028; the BMS collaboration expands in 2027.
How far are these results from actual market approval?
Per Anthropic's blog, human clinical trials could start as early as late 2028, while the full path from clinical trials to approval has no disclosed timeline. The blog frames the work as early progress, with all programs in preclinical or early stages.