Scientist Questions if Anthropic Claude Truly Made Biological Discovery on Its Own

Anthropic Claude Faces Scientific Scrutiny Over Claimed AI Discovery

Anthropic's announcement last week that its AI model discovered new enzymes with biotechnology promise has sparked controversy after computational biologist Mario Rodríguez Mestre revealed he and his colleagues had been studying the same enzymes for four years1

1

. The claim that Anthropic Claude made a biological discovery on its own now faces questions about whether the AI truly reasoned independently or was guided by previously shared research.

Dr. Mestre stated that for the past three years, his team regularly used Anthropic's AI models for writing code, drafting manuscripts, and performing other tasks, during which they shared key findings about the enzymes with Anthropic Claude1

1

. "So, for me, the most important question is not 'Were ARTs already known?' They were," Dr. Mestre said, raising concerns about whether this information was used to train future versions of the models1

1

.

The Disputed Biological Discovery and Array-Associated Reverse Transcriptases

Anthropic researchers claimed they instructed AI agents to search databases containing billions of gene sequences, tasking the AI model with finding genes that direct production of enzymes called reverse transcriptases1

1

. Nearly 950 Claude agents processed 210 million tokens over 21 hours before one flagged an unusual repeating pattern2

2

. The agents identified genes for possible reverse transcriptases in bacteria-infecting viruses known as jumbo phages, with each sequence located close to an array of genes for RNA molecules1

1

.

The Anthropic scientists named these enzymes array-associated reverse transcriptases, or ARTs, claiming their AI had taken "the first steps of a biological discovery on its own"1

1

. However, Dr. Mestre disputed this claim, stating "They are the same systems we have been studying for years," and that Anthropic Claude identified the same enzymes he had described to the AI1

1

.

Prior Research and Patent Filing Complicate AI Discovery Claims

Dr. Mestre first encountered the enzymes in jumbo phages in 2022 while searching for new reverse transcriptases1

1

. He worked briefly as a consultant for ReNegade Therapeutics, which filed a patent in 2023 covering several reverse transcriptases, including those he discovered, with Dr. Mestre listed as a co-inventor1

1

2

2

. After working for ReNegade, Dr. Mestre joined the University of Copenhagen to study reverse transcriptases in more detail, including the ones he found in bacteriophages, which his team nicknamed "jumbotrons"1

1

.

Anthropic released Claude in 2023, a year after Dr. Mestre began researching jumbotrons, and he found it useful for various tasks, stating "I completely switched my way of doing science"1

1

. Dr. Mestre said that he and his colleagues made the key discovery of identifying an array of RNA molecules and a protein associated with reverse transcriptase in a system over a year ago, which Anthropic claims was Claude's primary contribution1

1

.

Anthropic's Response and Broader Implications for AI Capabilities in Scientific Research

In a statement on Sunday, Anthropic said: "We are not aware of any previously published work describing the ART system we recently found. Claude was also not trained on any user transcripts, and our molecular biology team has no such access, either"1

1

. This controversy mirrors similar concerns that surfaced recently when researchers at OpenAI claimed to have solved a longstanding mathematics puzzle, after a leading mathematician noted he had been using the company's AI to work toward a solution for some time1

1

.

Despite the controversy, Feng Zhang, one of the pioneers of CRISPR genome editing, called the discovery "genuinely intriguing" and said it merits further investigation2

2

. Anthropic, expected to go public at a valuation potentially as high as $2 trillion, has touted biological research as central to its mission1

1

. CEO Dario Amodei stated at the United Nations that "we believe this type of application will accelerate rapidly in the years ahead, and that we can make profound progress in the near future on diseases that have plagued humanity for millennia"1

1

.

Anthropic's Expanding Life Sciences Push and AI-Driven Developments

Anthropic has been aggressively expanding its AI-driven developments in life sciences. Earlier this month, the company established a physical wet lab in the San Francisco Bay Area, expanding its AI-driven life sciences research into hands-on experimentation2

2

. Life sciences leader Eric Kauderer-Abrams emphasized that real-world lab work remains essential for biology, while the company combines internal research with external collaborations2

2

.

In the same week, Anthropic announced a collaboration with Novo to accelerate the development of new medicines, with Novo using Anthropic's advanced AI models to enhance software development and expand AI adoption2

2

. In July, Anthropic launched Claude Science, an AI workbench designed for scientists, supporting fields including genomics, proteomics, structural biology and cheminformatics2

2

. The platform integrates more than 60 scientific databases and tools and connects with Nvidia's BioNeMo Agent Toolkit for access to specialized biology models2

2

.

This controversy raises fundamental questions about AI safety, transparency in AI making a biological discovery on its own, and how companies should credit prior research when AI systems interact with existing scientific work. Watch for potential policy discussions around disclosure requirements when AI companies claim scientific breakthroughs, especially as competition intensifies in the gene databases analysis space. The outcome could influence investor confidence as Anthropic approaches its anticipated public offering and shape how other AI companies approach claims about autonomous scientific reasoning.