Anthropic said Wednesday that Claude identified a previously uncharacterized enzyme system after searching more than 200,000 reverse transcriptases and spotting a CRISPR-like DNA repeat pattern.
Dubbed array-associated reverse transcriptases (ART), Anthropic said human researchers validated Claude's finding in the lab, though its function and potential biotechnology applications remain unknown.
Dario Amodei framed the work as an example of AI moving from assisting scientists toward driving fundamental biological discoveries, arguing that humans can increasingly use AI to generate hypotheses and then rapidly test and iterate on them experimentally.
"I'm proud of the resources Anthropic has invested in accelerating the public benefits of AI through the life sciences, and we're aiming both to grow our life sciences team and to work with other scientists to extend this approach to a broad range of problems," Amodei wrote in a post on X.
He said Anthropic sees this as an early step toward accelerating the broader biomedical pipeline, while stressing that fully autonomous laboratory experimentation is not yet being done and that the current lab work is performed by humans.
"It’s easy to dismiss this as a one-off or curiosity, but we’ve repeatedly seen a pattern where AI performance in new intellectual domains goes from weak to superhuman in a matter of a few years," Amodei added.
Claude's broader role in biology
Separately on Wednesday, Anthropic released a report on how Claude is being used by the WHO Regional Office for Africa, the Institut National de Recherche Biomédicale and other partners to accelerate the response to a deadly outbreak of Bundibugyo ebolavirus in the Democratic Republic of Congo.
Claude is helping turn health workers’ field reports into daily situation reports, analyze outbreak data and run disease models, support vaccine research and process viral genomic data. It cuts some tasks that previously took a full day to less than an hour, while human experts remain responsible for scientific and public-health decisions.
Anthropic set up a Bay Area wet lab for physical biology experiments, Reuters reported Friday. Kauderer-Abrams, Anthropic’s head of life sciences, said the firm conducts biology experiments internally and with partners
"We believe that to do biology, the final test is still and will be for a while in real lab work," Kauderer-Abrams said. "We absolutely are doing that today."
This comes shortly after Anthropic warned AI models could be misused to facilitate biological threats.
Anthropic said a researcher outside the U.S. used Claude in early-stage bird flu research, while another user researched grants and genetic changes to a mosquito-borne virus. The firm said it banned accounts after identifying attempts to bypass safeguards.
"We’re publishing this work because we believe we have a responsibility to disclose malicious misuse of our services," the company said earlier this month. "As models become increasingly capable, their risks will increase, unless AI developers and society’s defenders act to make them safer."