Anthropic bought a one-person biotech that ran 11 labs with Claude, and now has its own drug programs
Mithril's muscular dystrophy work is now in-house at Anthropic, which also shared early results from six partners. It's still mostly preclinical.
The skim
- Anthropic has bought Mithril Medicine, a one-person biotech whose founder used Claude to run experiments across 11 outside labs on treatments for rare muscular dystrophies. Anthropic acquired it in July and disclosed the deal Tuesday; terms weren't shared.
- Mithril's programs are now Anthropic's first in-house drug programs. Anthropic says a candidate could enter human trials as soon as late 2028.
- It also published early results from six partners, including Bristol Myers Squibb, and stressed that almost all of the work is still preclinical.
What happened
In a post on its drug development work, Anthropic said Mithril founder Cody Tranbarger "used Claude to organize contracts, logistics, and experiments with 11 external labs." Mithril took a possible treatment for one form of muscular dystrophy from an idea to promising results in mice in less than 18 months. Tranbarger is now at Anthropic and plans to start several new programs.
The main target is LAMA2-related congenital muscular dystrophy, a rare childhood disease. In Mithril's technical report, which hasn't been peer reviewed, seven mice given one antibody design showed about twice the grip strength of four control mice, R&D World reports. R&D World also notes the limits: treatment wasn't randomized, the two groups were tested months apart, and the report itself calls for a blinded repeat. A second program, for alpha-dystroglycanopathy, aims for trials as early as 2029.
Claude wasn't flawless. The team reviewed six Claude errors, at least three of which reached outside labs in emails people had edited and sent, according to R&D World.
The partner results
Anthropic also shared work with six organizations. At Generate Biomedicines, scientists used Claude Code to redesign part of a CAR T-cell cancer therapy; one design was about 70% shorter and let immune cells make roughly 10 times as many receptors in lab tests. For Divergence Labs, Claude generated roughly 35,000 antibody designs for a food allergy treatment in about a month. At Enveda, Claude caught all 11 issues human reviewers found in 1,061 pages of manufacturing records, plus seven they missed, cutting review time by 89%. Bristol Myers Squibb is using Claude to assess clinical trial plans for an early-stage drug.
Why it matters
AI labs have mostly pitched drug discovery as designing molecules faster. Anthropic is betting on something broader: that a model can carry much of the paperwork, coordination and analysis that makes drug development slow and expensive, to the point where one person can run a program. "There is no single problem to solve; the bottlenecks are distributed all throughout the drug development process," Anthropic wrote.
It's also a pitch to investors. Anthropic just filed to go public, and its filing named AI's potential to find new cures first among its biggest benefits, The Information reported, per The Next Web. Mithril follows the $400 million purchase of Coefficient Bio in April.
Who's affected: patients with rare diseases like these, and small biotechs that could run leaner. What's next: Mithril aims to pick formal drug candidates in 2027. Until then, these are promising lab results, not medicines.
Sources: Anthropic, R&D World and The Next Web.
