Anthropic Biolab Credits Swarm of Claude Agents With CRISPR-Like Enzyme Discovery
Anthropic announced that hundreds of its Claude AI agents autonomously identified a novel bacteriophage enzyme system during a 21-hour search across DNA sequences, marking the first milestone for the startup's wet lab.
AI Agents Unearth a Novel Enzyme System
Anthropic announced that its Claude artificial intelligence has autonomously uncovered a previously unknown enzyme system with mechanics reminiscent of the gene-editing tool CRISPR. The find represents the debut scientific breakthrough from Anthropic's newly established wet lab as the AI company gears up for a potential initial public offering.
Anthropic disclosed the finding early to showcase Claude’s utility in scientific research, though the company conceded that it is still deciphering the precise biological function of the enzyme and cannot yet determine whether it will yield practical or transformative applications.
Nearly 950 Claude Agents Mine 210 Million Tokens of DNA Data
The discovery emerged from an automated scan of a vast database containing genetic sequences. According to Anthropic, human researchers were responsible only for setting up the initial prompt and performing the subsequent physical laboratory experiments to validate the findings.
The analytical legwork was driven entirely by machine intelligence:
- Roughly 950 Claude agents operated simultaneously over the course of 21 hours.
- The system processed through 210 million tokens before a single agent identified an anomalous repeating pattern within the genetic data.
- Once the agent flagged the anomaly for human review, researchers ran follow-up biological analyses and wet-lab experiments.
The tests confirmed the existence of an uncharacterized enzyme system embedded inside bacteriophages—viruses known to infect and replicate within bacteria. Anthropic stated that sharing the preliminary results early helps provide transparency into its ongoing biomedical research and proves Claude's capacity to handle biological discovery workloads.
The Tech Race to Validate AI Through Hard Science
The laboratory milestone arrives as Anthropic attempts to draw top-tier researchers into its biological division and expand deeper into scientific domains, including AI-assisted drug discovery.
Leading artificial intelligence developers are increasingly looking to fundamental scientific research to validate the practical value and commercial viability of their advanced systems. While Anthropic focuses on wet-lab biology and genetics, rival OpenAI has pushed heavily into advanced mathematics, aggressively working through complex problems as frontier labs compete to demonstrate that their models can achieve genuine scientific breakthroughs.
