Stanford Trains AI on DNA, Produces 16 New Bacteriophages
Synopsis
The breakthrough shows how generative AI can design novel virus genomes for scientific research, with potential applications in medicine and biotechnology.
Researchers at Stanford University have used artificial intelligence to design 16 fully functional viruses capable of replicating in laboratory conditions, marking the first time AI has successfully generated complete viral genomes.
The viruses, known as bacteriophages, infect only bacteria and pose no threat to humans. Scientists say the achievement could accelerate the development of treatments for antibiotic-resistant infections while also expanding the role of AI in synthetic biology.
The work relied on AI models called Evo1 and Evo2 which were trained on genetic sequences from viruses, bacteria, plants and humans before being refined to design bacteriophages. Researchers synthesised 302 AI-generated viral designs in the laboratory with 16 proving effective against E. coli bacteria.
The findings, published in the journal Science have drawn praise for their scientific potential but also renewed debate over biosafety and the governance of increasingly capable AI systems.
A Step Forward for AI-Driven Drug Discovery
Unlike large language models that predict words, Evo1 and Evo2 analyse the language of DNA. That capability allowed researchers to generate complete viral genomes rather than individual genetic components.
Scientists believe the technology could help develop new bacteriophage therapies as antimicrobial resistance becomes a growing healthcare challenge.
According to the World Health Organization (WHO), antimicrobial resistance was associated with 4.71 million deaths globally in 2021. It illustrated the urgent need for alternative treatments beyond conventional antibiotics. AI-designed bacteriophages could eventually become one of several tools used to tackle drug-resistant bacterial infections.
Scientists Urge Caution alongside Innovation
The breakthrough has also prompted fresh debate over the risks of AI-generated biology. In an accompanying commentary in Science, researchers from the Johns Hopkins Center for Health Security warned that advances in generative viral design raise important biosafety and biosecurity questions.
They argued the challenge is no longer whether such technology will exist but how it can be developed responsibly without increasing the risk of misuse.
The Stanford team said several safeguards were built into the project. Models were trained to design viruses that infect bacteria rather than humans. The team added that potentially dangerous viruses were excluded from the training data and all experiments were conducted in secure laboratory environments.
As AI continues to reshape scientific research, experts say technical progress will need to be matched by equally robust oversight and governance.
Source: BBC and Forbes
At Inspirepreneurs Magazine, covering entrepreneurship, business failures, and the human stories behind the world's most ambitious founders. She writes at the intersection of strategy and storytelling.
You Might Also Like
Australia’s Retail Sales: Stage 3 Tax Cuts and Wage Growth Effects
Anthropic Nears $20B Revenue Amid Pentagon Tensions
