AI Just Created 16 New Viruses From Scratch

Scientists have used artificial intelligence to generate new viruses from scratch that can infect bacteria and help overcome antibiotic resistance in laboratory tests, a breakthrough that researchers say could advance medicine but also raises biosafety concerns.

The study, published in the journal Science, describes how researchers from Stanford University and Arc Institute trained an AI model called Evo on genetic sequences from millions of organisms before using it to design thousands of new viral genomes.

Researchers synthesized and tested about 300 of the AI-generated genomes in the laboratory and found that 16 were viable viruses capable of infecting Escherichia coli (E. coli) bacteria. The viruses are bacteriophages, or phages, which infect bacteria but do not infect humans.

The researchers said the AI model learned the evolutionary constraints that shape natural genomes, but one of the newly created viruses contained a feature that was “evolutionarily distant,” suggesting the model generated changes that could have taken millions of years to emerge through natural evolution.

Laboratory tests showed that a cocktail of the new phages was able to overcome antibiotic resistance in some E. coli strains, while a comparable mixture of naturally occurring phages could not.

The findings “lays out a path for generating adaptive and resilient phage therapies against rapidly evolving pathogens,” the study authors wrote.

The work comes as scientists search for new ways to tackle antimicrobial resistance, which has made bacterial infections increasingly difficult to treat with existing antibiotics.

“The breakthrough achieved is significant,” Jordi García Ojalvo, a professor of systems biology at Pompeu Fabra University, said in comments provided through the Science Media Centre.

However, experts said the advance also highlights the need for stronger safeguards around AI tools capable of designing biological systems.

“Although this is promising for life sciences applications, it also raises urgent biosafety and biosecurity questions,” doctors from the Johns Hopkins Center for Health Security wrote in a commentary accompanying the study in Science.

“The ability to compose viral genomes using generative AI now exists; the governance to safely steer it does not,” they wrote.

The commentators said the Stanford-led team had addressed biosafety issues more thoroughly than many developers of biological AI models but warned that similar techniques should not be applied to pathogens that infect humans, animals or plants.

“Such genomes might encode new pathogens that can infect humans, animals, or plants in ways that cannot be contained by existing countermeasures,” they wrote.

The researchers stressed that their work focused on bacteriophages that infect E. coli and not viruses capable of infecting humans, and it remains unclear whether the approach can be extended to more complex viruses.

García Ojalvo said the immediate biosecurity risks were limited because each AI-designed genome must still be synthesized and tested individually, and the model produced only 16 viable viruses from hundreds of thousands of generated candidates.

“It is difficult to imagine these models automatically generating viable genomes ‘out-of-the-box,'” he said.

Stay ahead of the Stories shaping our world. Subscribe to Impact Newswire and join our 
WhatsApp Channel for updates on global tech, business, and innovation—all in one place.

Dive deeper into the future with the Cause Effect 4.0 Podcast, where we explore the ideas, trends, and technologies driving the global AI conversation.

Got a story to share? Contact Us to reach a global audience with Impact Newswire.