Anthropic disclosed on October 27, 2026, that it had acquired Mithril Medicine, a biotech whose programs became Anthropic’s first internal drug-development programs. The acquisition reportedly closed in July. A pilot mouse study tied to one program reported higher muscle force in treated animals, but it remains an exploratory preclinical result.
What Anthropic acquired
Mithril Medicine was founded by Cody Tranbarger, its sole full-time employee. Anthropic said its two muscular-dystrophy programs joined the company’s internal drug-development work, and Tranbarger joined Anthropic to continue the programs and help lead that effort. The reported July closing and the October public disclosure are separate milestones.
Claude coordinated the work; people and labs did the science
Claude helped organize contracts, logistics, and experiments across 11 external laboratories. Mithril also had eight paid consultants and advisers. People made the scientific decisions, while contract research organizations performed the physical experiments.
The technical report also describes errors in this workflow. Its authors examined six Claude errors: at least three reached contract research organizations in emails that people had edited and sent. Two errors in Claude’s review of production data went unnoticed for more than four days. The report says Claude compiled the account of its own work from its threads and notes, which the authors identify as a potential source of bias.
The mouse result concerns LAMA2-CMD
The reported animal study focused on LAMA2-related congenital muscular dystrophy, or LAMA2-CMD. The experimental bispecific antibody was designed to bridge laminin-α4 and β-dystroglycan, two proteins involved in connecting muscle fibers with surrounding tissue. Trade reporting identifies alpha-dystroglycanopathy as Mithril’s other program.
In the pilot, seven dyW mice received the 2+2 IgG-scFv antibody and four mice received a vehicle control. The treated animals received 100 mg/kg by intraperitoneal injection twice a week from birth. The technical report recorded peak triceps force at 1.95 times the vehicle group’s mean (95% CI 1.51–2.51). Its litter-level comparison yielded P = 0.057.
That is an exploratory result in mice, not a patient outcome. The technical report was not peer-reviewed, and the study was neither randomized nor blinded. The treated mice were tested in sessions 126 and 161 days after the vehicle sessions, with a colony move in between. The authors call for a concurrent, blinded replication.
Anthropic said on October 27 that a drug could be ready to enter human clinical trials as soon as late 2028 if its candidates continue to hold up. That is a conditional projection; the reported mouse study is an earlier stage of development.
