Mithril Medicine had one full-time person and a drug candidate for a rare form of muscular dystrophy. On October 27, 2026, Anthropic said it had acquired the company. Its founder, Cody Tranbarger, used Claude to handle contracts, logistics, and experiments with 11 outside labs, and handed many other parts of running the business to it as well. Mithril took a possible treatment from an idea to promising results in mice in less than 18 months.
Tranbarger now works at Anthropic, where he's continuing the research and plans to start several new programs. Mithril was working on treatments for two rare forms of muscular dystrophy. Anthropic says that if the drug candidates keep holding up, one could be ready to start human trials as soon as late 2028. PYMNTS noted that no price was disclosed.
For anyone following the one-person-company idea, this is one of the clearest examples yet. A single founder ran a biotech, the kind of business that usually needs a staff of scientists, project managers, and lawyers, by having Claude coordinate the work across labs he didn't own. We've written about the software version of that shift in what happens when everyone makes their own software.
What Claude did for six drug companies
The purchase came inside a longer update on drug work Anthropic has been doing with six partners. Its argument is that there isn't one big bottleneck in making a drug. The slow parts are spread across research, design, manufacturing, paperwork, and trials, so Claude has to help with all of them.
At Generate Biomedicines, which is trying to make CAR T-cell therapy work on solid tumors, six scientists used Claude Code to propose new designs for one piece of the therapy's genetic instructions. One design was about 70% shorter than the standard version and worked better in lab tests, letting immune cells make roughly 10 times as many of the receptors they use to spot cancer.
Divergence Labs is working on food allergies. Claude turned its goals into a design plan and then, largely on its own, generated about 35,000 antibody designs in roughly a month. Early tests found three designs that bind to IgE, the molecule behind allergic reactions, and 20 that pull it off immune cells. Three of those 20 did that better than the standard medication.
BigHat Biosciences had its lab make and test about 1,010 variants of two drug candidates in about 10 weeks, based on changes Claude proposed. Partway through, the team suspected the standard stability and clumping measurements were misleading, so Claude and an Anthropic researcher reanalyzed the raw data and proposed better ones. The best versions bound their targets more than 10 times more tightly than the originals.
Enveda's manufacturing chemists spent up to 75% of their time on paperwork, so it connected its manufacturing records to Claude and had Claude flag problems. On 1,061 pages, Claude caught all 11 issues human reviewers had found plus seven they'd missed, and review time fell 89%, from more than 25 hours to under three. Claude also drafted one regulatory document in four hours that took about three months by hand.
The Column Group is building a new company, Pleiades, meant to run more drug programs with Claude than a small team normally could. Bristol Myers Squibb is using Claude to plan early clinical trials and estimate their odds. Its head of early clinical development for hematology, oncology, and cell therapy, Michael Pourdehnad, said Claude produced an initial analysis "from a single prompt" that would otherwise have taken days.
Anthropic also has a business reason to make this case. Its amended IPO filing says its customers include nine of the 10 largest pharmaceutical companies, and that it bought a computational biology company, Coefficient Bio, in April 2026. OpenAI is chasing the same market. PYMNTS reported that it agreed on October 20 to acquire Orchestra Bio. Drug companies spend heavily, and a credible story about Claude cutting years off drug development is worth a lot to a company about to go public.
The Enveda paperwork result and the Mithril timeline are the two to watch. One shows Claude saving real hours on work nobody wants to do, and the other will show, by late 2028, whether a drug run mostly by one person and an AI makes it into a human trial.
Sources
Anthropic: Early progress on accelerating drug development with Claude - October 27, 2026. The Mithril acquisition, Tranbarger's use of Claude, the 18-month timeline, the late-2028 trial estimate, the six partner results, the Pourdehnad quote, and the preclinical caveat.
PYMNTS: Anthropic buys one-person biotech to prove Claude can make drugs - November 2, 2026. No disclosed price and OpenAI's agreement to acquire Orchestra Bio.
SEC: Anthropic, PBC Form S-1/A - Filed November 2, 2026. Nine of the 10 largest pharmaceutical companies as customers and the April 2026 Coefficient Bio acquisition.
