Anthropic Is Developing Its Own Drugs, and One Founder Ran 11 Labs With Claude

An AI company walks into the lab

Anthropic is no longer only selling the tool. The Rundown AI reported on October 28 that the company shared early results from its drug development partners and revealed it is now running its own drug discovery programs, including buying Mithril Medicine, a biotech run by a single founder.

The headline inside the headline is the shape of that company. Mithril founder Cody Tranbarger built a muscular dystrophy drug candidate with Claude helping him juggle contracts and tests at 11 outside labs. The candidate showed promise in mice within 18 months and could reach human trials as early as late 2028.

A pipeline in a person

Tranbarger also turned out to be Professor Oak, the name behind a biotech account on X run under a pseudonym for six years. His setup has been described as a pipeline in a person: one operator, one AI assistant, and a network of outside labs doing the bench work he coordinates.

The Neuron confirmed the same shape from its side, reporting that Anthropic acquired a one person drug development company that used Claude to coordinate 11 labs and reported early promise in mice. Two newsletters, same picture: the laboratory is a network, and the founder is the control room.

What Claude did at the partner companies

The partner results show where the hours went. At Enveda, Claude wrote a first draft of a key regulatory filing in 4 hours, work that took roughly 3 months by hand, and it also flagged seven batch record issues that humans had missed.

At Divergence Labs, three of roughly 35,000 antibodies Claude designed appeared to beat the current main drug at unhooking allergy triggers from cells. Designing candidates and drafting filings are different kinds of work. Claude is showing up in both.

The money context, stated honestly

Anthropic is not hiding why this market tempts it. Its recent IPO filing pitched a global pharmaceutical market expected to top 2 trillion dollars this year. The same filing warned that its own drug bets may never generate revenue.

Hold both facts at once. The upside is one of the largest markets on earth. The risk is real enough to print in a filing. That honesty is part of why the story matters more than the hype around it.

The operator lesson for small teams

Strip away the lab coats and this is a story about leverage. One founder did not hire a building full of scientists. He coordinated specialists he did not employ, with AI carrying the contracts, the scheduling, and the paper trail between them.

That is the same shape as any small creative business that punches above its size. The owner directs. Systems carry the coordination. Specialists plug in per project. The scarce skill is no longer doing every task. It is running the network without dropping a ball.

• Coordinate specialists per project instead of hiring every skill full time
• Let AI carry contracts, drafts, and records while you carry decisions
• Use AI as a second checker, the way the Enveda filing review caught missed issues
• Say the risk out loud in your own offers, the way the filing admits the bets may fail

One person, wider reach

A solo founder reached the edge of human trials in a field that used to require an institution. The tool did not replace the labs. It made one person able to run them.

Whatever your field is, that ratio is the story to study.