AI Designs 16 Functional Viruses From Scratch in Lab First

AI crosses new threshold in biological design

Researchers at Stanford University have successfully used artificial intelligence to design 16 fully functional viruses from scratch—the first time complete genomes have been created by AI and proven viable in laboratory conditions.

The team, led by assistant professor Brian Hie, employed AI models called Evo1 and Evo2 to generate novel bacteriophages, viruses that infect only bacteria and pose no threat to humans. Out of 302 AI-generated designs synthesized in the lab, 16 successfully killed E. coli bacteria, according to findings first reported by the BBC.

"This is a next step in the complexity that's designable by generative AI, this is the first time generative AI has been used to design a complete genome," Hie told the BBC. The technology works similarly to large language models like ChatGPT, but instead of predicting sequences of text, these models predict genetic sequences—what Hie calls "the language of life."

The AI models were trained on genetic codes from viruses, bacteria, plants, and humans, then refined to produce bacteriophages targeting specific bacterial species. PhD student Samuel King described the moment of discovery in the early morning hours when clear spots appeared on petri dishes, indicating the engineered phages were successfully consuming bacteria. When results were shared with the team, "the room spontaneously burst into applause," Hie recalled.

Why it matters

This breakthrough represents a fundamental shift from analyzing existing biology to designing new biology computationally. The ability to engineer bacteriophages could provide critical new weapons against antibiotic-resistant infections, while the broader capability to design synthetic genomes opens possibilities for custom enzymes, therapeutic antibodies, and treatments for genetic disorders. However, the same technology that could accelerate drug development also creates pathways for designing harmful pathogens—a dual-use dilemma that demands immediate policy attention.

Safety concerns prompt urgent questions

In a commentary accompanying the study's publication in the journal Science, researchers from Johns Hopkins University's Center for Health Security raised what they termed "urgent biosafety and biosecurity questions." Dr. Thomas Inglesby and Dr. Moritz Hanke wrote that the issue is no longer whether generative viral genome design will exist, but whether it can be deployed without "enabling serious harm."

The Stanford team implemented multiple safety measures: they excluded viruses capable of infecting complex organisms from their training data, focused exclusively on bacteriophages rather than human-infecting viruses, and conducted all work in secure laboratory facilities. Hie maintains that existing safeguards provide substantial protection for ensuring the technology serves beneficial purposes.

What comes next

The bacteriophages created in this study have genomes approximately 5,400 base pairs long. By comparison, the simplest living cells contain roughly 500,000 base pairs, while human genomes span three billion base pairs. Hie acknowledged that designing simple living organisms "would probably be a lot of work, but not impossible," and expressed interest in pursuing that direction.

Prof. Marc Güell from Pompeu Fabra University in Spain characterized the research as a "very significant turning point" because "for the first time in history, we are beginning to design biology on a computer." Prof. Patrick Cai from the Manchester Institute of Biotechnology called it an "important milestone" suggesting that AI models are beginning to learn evolutionary design principles, "opening the door to AI-assisted genome writing."

These details were first reported by the BBC.