Artificial intelligence is moving deeper into biology, and Anthropic has just revealed an early example of what that could mean.
The company says its Claude AI model helped identify a previously undescribed enzyme system with characteristics reminiscent of CRISPR. The discovery is the first major result announced by Anthropic’s new life-sciences research effort. The company’s announcement was published on September 23, 2026. Anthropic’s research announcement
What Did Claude Actually Find?
Anthropic calls the system array-associated reverse transcriptases, or ART.
The system was found in bacteriophage DNA. Bacteriophages are viruses that infect bacteria. Anthropic says the system contains a reverse transcriptase, a neighboring protein and a sequence of repeated DNA elements.
That arrangement caught Claude’s attention because the repeated sequences resemble the arrays associated with CRISPR systems.
According to the company, nearly 950 Claude agents searched DNA data for about 21 hours. They examined roughly 200,000 enzyme clusters before narrowing the candidates for human researchers to investigate.
Why Scientists Are Interested
CRISPR became one of modern biology’s most important technologies because researchers learned how to adapt a natural bacterial defense mechanism for programmable gene editing.
The National Human Genome Research Institute explains how CRISPR-Cas systems can be used to target specific DNA sequences.
The new ART system is not yet a replacement for CRISPR.
That distinction is important. Anthropic says it does not yet know what ART actually does. Early experiments have identified intriguing biological features, but the system still requires substantial laboratory investigation.
The AI Discovery Pipeline Is the Bigger Story
The significance may extend beyond this particular enzyme.
Anthropic created its life-sciences research group in spring 2026 to investigate whether AI agents could accelerate biological discovery. The company combines computational analysis with experiments conducted by human scientists in its laboratory.
The National Institutes of Health has supported AI and computational approaches across biomedical research, reflecting the broader movement toward using machine learning to analyze enormous biological datasets.
AI can search patterns across biological information at a scale that would be difficult for individual researchers to reproduce manually.
That does not mean AI replaces scientists. Instead, it could change how researchers identify promising questions.
Why Biosecurity Experts Are Watching
The same capability that could accelerate medical research also raises questions about responsible use.
AI systems capable of analyzing biological information could potentially help researchers identify useful mechanisms faster. But researchers and governments also have to consider how increasingly capable AI could interact with sensitive biological knowledge.
The World Health Organization has emphasized the importance of biosafety and responsible management of biological risks.
Anthropic says its current laboratory work is performed by human scientists and that the facility does not handle pathogens capable of infecting humans.
Discovery Is Not the Same as a New Gene-Editing Tool
The excitement surrounding the announcement needs context.
Feng Zhang, a pioneer of CRISPR research, described the identification of the RNA-repeat arrays as intriguing and worthy of further investigation. Other scientists have cautioned against immediately describing ART as a “new CRISPR.”
The system’s biological function remains unknown, and Anthropic’s results have been released as a preprint rather than a fully peer-reviewed study.
That makes the discovery an important research lead rather than an established medical technology.
A New Phase of the AI Arms Race
The competition in artificial intelligence is increasingly moving beyond chatbots and computer code.
Companies are now exploring AI for drug discovery, molecular biology and scientific research. The ability to find previously overlooked biological patterns could become an important advantage in the next generation of technology.
Anthropic’s enzyme discovery does not prove that AI has created the next CRISPR.
It does show something potentially more significant: AI agents are beginning to participate in the process of finding biological questions humans can investigate.
The next challenge will be determining what these discoveries actually do — and ensuring that the growing power of AI-driven biology develops alongside equally strong scientific oversight and biosecurity safeguards.
#Anthropic #ClaudeAI #CRISPR #ArtificialIntelligence #AIResearch #Biotechnology #GeneEditing #Biosecurity #Science #Tech