# Anthropic AI discovery of ART enzymatic system > Live situation record from CLSTR: https://clstr.news/situations/anthropic-ai-discovery-of-art-enzymatic-system > Updated: 2026-09-28T16:46:47.000Z. Sources: 96. Developments: 3. Anthropic has announced the discovery of a new biological system, named Array-Associated Reverse Transcriptases (ART), using its Claude AI models. The discovery was made through an automated search conducted by approximately 950 Claude-based agents over a 21-hour period, during which they analyzed over 200,000 reverse transcriptases within massive genetic databases. Found within the DNA of bacteriophages, the ART system consists of a reverse transcriptase enzyme, an adjacent gene, and a long array of regularly spaced DNA repeats. While the structure bears a resemblance to the CRISPR gene-editing mechanism, researchers noted there is currently no evidence that ART functions as an advanced CRISPR system or can be used for gene editing. Dario Amodei, co-founder and CEO of Anthropic, noted that the AI identified the pattern in genomic data and proposed experiments for human scientists to validate. Laboratory testing has confirmed that the repeating sequences in the ART system transcribe into RNA, producing various short RNA molecules. However, the specific biological function of these molecules and the broader role of the ART system within the bacteriophage remain under investigation. Recent claims regarding the novelty of the discovery have been disputed by Mario Rodríguez Mestre, a computational biologist at the University of Copenhagen. Mestre asserts that he and his colleagues have been studying these enzymes, which they refer to as ‘jumbotrons’, since 2022. He further claims his team had shared unpublished research details through Anthropic’s Claude models, raising questions about the independence of the AI’s findings. While some experts, including MIT professor Feng Zhang, view the identification as an exciting example of AI’s potential in biology, the dispute highlights tensions between human research and machine learning outputs. ## Claims - Anthropic's AI agents independently identified a previously unknown biological system called array-associated reverse transcriptases (ARTs). (corroborated by 3 sources) - The AI process involved 950 Claude agents processing 210 million tokens over 21 hours to identify the enzyme system. (corroborated by 2 sources) - Computational biologist Mario Rodríguez Mestre claims he and his team identified these enzymes in 2022. (corroborated by 2 sources) - ReNAgade Therapeutics filed a 2023 patent covering several reverse transcriptases, including those identified by Mestre. (corroborated by 2 sources) - Mestre asserts that his team had shared unpublished details of their research regarding these enzymes through Anthropic’s Claude models. (single source) - MIT professor Feng Zhang described the identification of the RNA-repeat arrays as an exciting example of AI's contribution to biological discovery. (single source) ## Timeline ### 2026-09-28: Anthropic AI discovery disputed by researcher A researcher is challenging Anthropic's claim that its Claude AI independently discovered new gene-editing enzymes, asserting he had documented the same biological system years earlier. 3 sources. https://clstr.news/cluster/anthropic-ai-discovery-disputed-by-researcher ### 2026-09-26: Anthropic AI discovers new biological system in bacteriophage DNA Anthropic's Claude AI agents discovered a new biological system, named ART, in bacteriophage DNA, featuring a structure similar to CRISPR after analyzing massive genomic datasets. 4 sources. https://clstr.news/cluster/anthropic-ai-discovers-unknown-biological-system-in-bacteriophage-dna ### 2026-09-23: Anthropic discovers new ART enzymatic system using Claude AI Anthropic's Claude AI agents identified a new enzymatic system called ART in bacteriophage DNA, discovering the mechanism in just 21 hours by analyzing over 200,000 reverse transcriptases. 90 sources. https://clstr.news/cluster/anthropics-claude-ai-discovers-new-art-enzymatic-system --- Cite as: Anthropic AI discovery of ART enzymatic system. CLSTR, https://clstr.news/situations/anthropic-ai-discovery-of-art-enzymatic-system