Artificial intelligence has reached a major new milestone in fundamental research. By deploying a swarm of 950 Claude agents working autonomously, Anthropic has successfully identified a novel enzymatic system within the genetic material of bacteriophages. This project, conducted without direct human assistance, utilized approximately 210 million tokens over a 21-hour period. This technological feat allowed the system to sift through a vast database to isolate 3,500 promising candidates, of which about 20 underwent in-depth analysis—a task that would have required months or even years for a traditional team of researchers.
The discovered system, dubbed ART (array-associated reverse transcriptases), is particularly intriguing to the scientific community due to its structure. Its genetic layout bears striking similarities to the CRISPR system, notably in the presence of tandem repeat sequences and a partner protein whose biological function remains a mystery for now. Given that genetic systems with such a configuration are often capable of manipulating DNA through cutting or copying processes, this finding could pave the way for new programmable technologies in the biotechnology sector.
The involvement of renowned experts, such as Feng Zhang from MIT, underscores the seriousness and interest surrounding these results. This discovery is part of a trend where AI is no longer merely analyzing existing data, but is instead generating original research hypotheses. Anthropic, however, emphasizes the complementarity between machine and human: while Claude identifies targets, rigorous physical testing in a biosafety laboratory remains the only way to validate the actual properties of these enzymes in a controlled environment.
This breakthrough confirms the disruptive potential of AI models in complex fields like genomics. After demonstrating its ability to uncover critical vulnerabilities in computer code, AI is now establishing itself as a catalyst for biological discovery. By automating the screening of vast sequence libraries, this tool radically changes research methodology, shifting science from empirical observation to an industrial acceleration of discovery. The scientific community is now invited to collaborate with these systems to explore the mysteries of life with unprecedented efficiency.