New AI-Discovered Enzyme System May Reshape Genetic Toolkits

Claude Has Discovered a Previously Unknown Enzyme System Hidden in Bacteriophage DNA

September 23, 2026 — Anthropic announced Wednesday the first result from a new life sciences research group, established by the company in spring 2026: its Claude AI model autonomously identified a previously uncharacterized enzyme system hidden in the DNA of bacteriophages, the viruses that infect bacteria. The company shared the finding through a blog post and a preprint, which has not yet undergone peer review.

What Claude found

The system, which the team named array-associated reverse transcriptases (ART), consists of three parts: a reverse transcriptase enzyme (RT, which copies RNA into DNA), a neighboring partner gene, and a long array of evenly spaced DNA repeats. According to the preprint, these arrays contain between 3 and 21 copies of a short repeat — a layout that visually resembles CRISPR arrays, although the system lacks the cas genes typically found alongside CRISPR.

In lab testing, Anthropic’s scientists found that the array is transcribed into a set of distinct short RNAs; in previously published data from a Staphylococcus phage, these RNAs accounted for as much as 8% of the phage’s total RNA 15 minutes after infection. Importantly, the preprint does not yet show whether the enzyme is active or whether it actually acts on these RNAs — so this remains a research hypothesis rather than a confirmed function.

Anthropic notes that only a small number of known biological systems share this particular combination of features, and all of them are capable of cutting, copying, and inserting genetic material — the kind of programmable capability that, in CRISPR’s case, helped revolutionize genetic medicine.

How Claude did the work

The team gave the model a single research brief: search a database of 1.9 billion protein clusters for new reverse transcriptase systems. Roughly 950 Claude agents ran for about 21 hours, using 210 million tokens, to scan over 200,000 known reverse transcriptases, narrowing the search to 3,500 candidates and finally to 20 systems for detailed analysis.

According to reports, while reading the raw DNA sequence next to one of these enzymes, the agent flagged what it recognized as a CRISPR-like repeat pattern — then counted the repeats, measured their spacing, compared the layout to known RT systems, and checked the scientific literature to see whether the pattern had been previously reported, before filing a report for human review. All actual laboratory experiments were carried out by human scientists; Claude’s role was limited to searching genomic datasets, generating hypotheses, analyzing candidates, and helping interpret results.

Reactions from the scientific community

Feng Zhang, one of the pioneers of CRISPR gene editing and a professor at MIT and the Broad Institute, reviewed the preprint and called the finding an exciting example of how AI agents can contribute to biological discovery, adding that the identification of these RNA-repeat arrays associated with reverse transcriptases is intriguing and merits further investigation.

Not all reactions were as enthusiastic: researcher Mark Riedl publicly expressed skepticism, suggesting it isn’t yet clear whether Anthropic has actually discovered something functionally significant, given that the system’s function remains unknown.

The takeaway

This is a preliminary finding, not yet published in peer-reviewed form, and Anthropic itself acknowledges that understanding ART’s actual function will take considerably more work. The company says it chose to share the discovery at this stage both to demonstrate Claude’s capabilities and to give the scientific community visibility into the direction of its research.


Analysis produced with Claude (Anthropic), with editorial approval and oversight from Robert Williams. Readers are encouraged to consult additional sources; this piece does not represent the views of the parties quoted and should be read as journalistic reporting, not as a statement on their behalf.

Source: https://x.com/AnthropicAI/status/2102824959827742916?s=20


Discover more from #News247WorldPress

Subscribe to get the latest posts sent to your email.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Discover more from #News247WorldPress

Subscribe now to keep reading and get access to the full archive.

Continue reading

Discover more from #News247WorldPress

Subscribe now to keep reading and get access to the full archive.

Continue reading