AI Designs First Functional Virus Genomes from Scratch

Researchers from the Arc Institute and Stanford University have used generative AI to design entirely new, functional viruses from scratch. The breakthrough marks the first time artificial intelligence has been used to create a complete, viable genome.

Using genome language models known as Evo 1 and Evo 2, the team developed a series of synthetic viruses. These models were trained on trillions of nucleotides to master the complex "grammar" of genetic code. After the AI generated a vast pool of candidates, researchers synthesized 302 of the most promising designs in the lab. Of those, 16 novel viruses were proven to successfully infect and replicate within E. coli bacteria, according to the study published in the journal Science.

The Shift to Biological LLMs

This project adapts the architecture of Large Language Models (LLMs)—the technology powering tools like ChatGPT—and applies it to biological sequences rather than human text. While AI has previously assisted in the design of specific components like antibiotics, the creation of a full, working genome is significantly more complex. To mitigate risks during development, the researchers implemented safety guardrails by excluding viruses that infect humans from the training data and conducting all experiments in secure laboratories.

Implications for Medicine and Security

The designed viruses are bacteriophages, which specifically target bacteria and are incapable of infecting humans, animals, plants, or fungi. This capability represents a turning point in synthetic biology, offering a path toward highly precise phage therapy to treat antibiotic-resistant infections and advancing the potential of gene therapy. Brian Hie, an assistant professor at Stanford University, noted that this represents a "next step in the complexity that's designable by generative AI."

However, the ability to program life from a computer screen introduces severe risks. Experts warn that the same technology used to create beneficial phages could be weaponized to engineer new, deadly diseases if the process is not strictly regulated. Dr. Thomas Inglesby and Dr. Moritz Hanke of the Center for Health Security at Johns Hopkins University stated that the findings raise "urgent biosafety and biosecurity questions."

The Path Forward

As the field moves toward the computer-aided design of biology, the scientific community must now balance innovation with oversight. Future research will likely focus on expanding the range of functional genomes AI can produce, while policymakers face the immediate task of establishing frameworks to prevent the misuse of generative biological models. For now, the success of the 16 synthetic viruses serves as a proof of concept that the fundamental code of life can be written by an algorithm.