// NATURE NEWS — SPAZIO & SCIENZA
Anthropic bought a one-person, AI-powered biotech: is this drug discovery’s future?
Heidi Ledford reports for Nature from Seattle, Washington.
Search author on:
PubMed
Google Scholar
Minimalist biotechnology companies that have just a handful of employees — or even as few as one — are becoming more feasible with the rise of artificial-intelligence tools.Credit: Getty
How many scientists does it take to discover a drug?
The question came to the fore after news that a biotechnology company armed with an artificial-intelligence assistant, two promising drug leads and just one full-time human employee had been acquired by San Francisco-based Anthropic as part of the AI giant’s expansion into biomedical science.
AI agents run this automated biology lab that sequences DNA and more
In truth, it took many human scientists to develop the molecules that are the mini-company’s star products: two antibodies that are still in the early stages of development and were designed to treat a rare form of muscular dystrophy. The company, Mithril Medicine in Cambridge, Massachusetts, also hired part-time consultants to advise on the project, and paid at least five contract research organizations to conduct experiments. Many more researchers will be needed to carry the drug further, especially if it eventually enters clinical trials. “It takes a village to develop a drug,” says Jeb Keiper, a partner at Third Rock Ventures, a venture-capital firm in Boston, Massachusetts.
But the news sparked a wave of discussions about whether the one-person biotech is the drug-discovery business model of the future. “We will see more and more of these small teams that are enabled by AI,” says James Zou, a computer scientist who works on biomedical applications of AI at Stanford University in California. “They will be very powerful for identifying good opportunities and moving quickly.”
Itsy-bitsy biotechs are not a new concept. Entrepreneur Matt Davidson started his first biotech company in 2013, nearly a decade before ChatGPT’s public debut. Davidson kept the roster trim for years, with only two full-time employees until the drug he was developing was ready to enter late-stage clinical testing. The company never had a physical laboratory, Davidson says. All the lab work was contracted out.
His current company, Azora Therapeutics in Los Angeles, California, has a whopping five employees. He uses AI more for summarizing reports than for scientific discovery and says that AI tools have made tiny biotechs more efficient. “We use consultants a little bit less than we used to,” he says.