According to Anthropic, Mithril advanced a potential treatment from an initial concept to encouraging results in mice in fewer than 18 months. Its research focuses on two rare forms of muscular dystrophy, including LAMA2-related congenital muscular dystrophy, using engineered antibodies designed to address the underlying disease mechanisms.
In a technical report that has not undergone peer review, Anthropic said mice receiving one experimental antibody showed approximately twice the grip strength of untreated animals. The findings remain preliminary, but the company believes one of Mithril's drug candidates could enter human clinical trials as early as late 2028 if further development is successful.
Founded approximately 18 months ago with seed funding from Atlas Venture, Mithril operated with Tranbarger as its sole full-time employee, supported by part-time consultants and external laboratories. Tranbarger previously worked in venture capital and crossover investing.
He describes the operating model as Pipeline-in-a-Person, in which one scientist uses AI to manage multiple drug development efforts simultaneously.
Tranbarger joined Anthropic to gain additional resources for Mithril's existing research, investigate how many drug programs an individual can oversee with AI assistance and explore whether the approach can be adopted more widely across biotechnology companies. Anthropic said he intends to initiate additional research programs.
The acquisition also marks a shift in Anthropic's involvement in pharmaceutical development. Beyond providing AI tools to outside researchers, the company has begun pursuing drug discovery internally, including work on rare childhood muscular dystrophies in which Claude performs substantial scientific and administrative tasks.
Anthropic's announcement included results from collaborations with BigHat Biosciences, Bristol Myers Squibb, Divergence Labs, Enveda, Generate Biomedicines and The Column Group. The projects illustrate how Claude is being used across drug design, laboratory research, regulatory documentation and clinical development planning.
At Generate Biomedicines, researchers used Claude Code to redesign a component of a CAR T-cell cancer therapy. One proposed design was approximately 70% shorter than the conventional version.
In laboratory testing, immune cells using the modified component produced roughly 10 times more cancer-recognizing receptors. Anthropic also reported improvements in tumor-cell killing and the duration of immune-cell activity.
Other collaborations focused on antibody engineering. Divergence Labs, which is developing treatments for food allergies, used Claude to produce approximately 35,000 antibody designs over one month.
Initial testing identified three designs that bound to IgE, an antibody involved in allergic reactions, along with 20 molecules capable of displacing IgE from immune cells. Three candidates appeared to outperform a standard medication in early experiments, according to Anthropic.
At BigHat Biosciences, researchers evaluated approximately 1,010 AI-proposed antibody variants over 10 weeks. The strongest candidates demonstrated binding affinity more than 10 times greater than the original molecules.
While these results suggest that AI can help researchers identify promising molecular designs, they are based on early laboratory experiments rather than evidence of effectiveness in patients.
Anthropic also highlighted applications outside molecular discovery, where administrative and regulatory requirements can consume substantial time.
Enveda tested Claude on 1,061 pages of pharmaceutical manufacturing documentation. The system identified all 11 issues previously found by human reviewers and detected seven additional problems they had not flagged.
The company reported an 89% reduction in human review time, bringing the process from more than 25 hours to fewer than three. In another demonstration, Claude prepared a regulatory document in four hours, compared with approximately three months for a comparable manual drafting process.
Bristol Myers Squibb is using Claude to examine clinical development plans for an experimental medicine and evaluate the likelihood of success in clinical trials. The companies intend to expand their collaboration in 2027.
Michael Pourdehnad, who leads early clinical development in hematology, oncology and cell therapy at Bristol Myers Squibb, said the technology substantially accelerated the initial assessment.
“From a single prompt, Claude generated a robust initial analysis,” Pourdehnad said.
The Column Group, meanwhile, is establishing a biotechnology company called Pleiades that will use Claude to coordinate an expanding portfolio of drug programs with a relatively small team. Its first specialized Claude tool, designed to evaluate potential drug targets, has been assessed against 45 combinations of diseases and targets.
These projects reflect Anthropic's effort to apply Claude across different stages of pharmaceutical development, rather than limiting its role to generating scientific hypotheses or analyzing research literature.
The Mithril deal is Anthropic's second life sciences acquisition this year. In April, the company purchased Coefficient Bio for approximately $400 million, adding another biotechnology-focused business to its operations.
Anthropic has also introduced Claude Science, a research environment for scientists, alongside AI usage credits for life sciences researchers and a verification program that provides eligible organizations with more flexible safeguards.
The company has previously reported using Claude to design functional protein-binding molecules, adding to its broader research into AI-assisted biological discovery.
Despite the reported progress, Anthropic acknowledges that its pharmaceutical research remains largely preclinical. The experimental treatments discussed in its announcement have not established safety or effectiveness in humans, and promising laboratory or animal results do not guarantee successful clinical development.
The company plans to expand its drug discovery efforts and increase Claude's involvement in clinical research as projects progress.
Mithril's acquisition gives Anthropic a direct role in developing experimental medicines while providing a test of whether AI can allow individual researchers and small teams to manage more complex scientific programs. The results reported so far demonstrate progress in laboratory design, documentation and research coordination, but the clinical potential of the resulting treatments remains to be determined.
This analysis is based on reporting from TNW.
Image courtesy of dealroom.
This article was generated with AI assistance and reviewed for accuracy and quality.
