AI-designed bacteriophage genomes raise biosecurity concerns
Researchers at Stanford University and the Arc Institute used AI to design functional bacteriophage genomes, including 16 that were synthesised and tested successfully in laboratories. The development raises biosecurity concerns about potential misuse of AI-designed viruses that could evade detection systems.
What happened
Researchers at Stanford University and the Arc Institute used generative AI models Evo 1 and Evo 2 to design complete genomes of bacteriophages—viruses that infect bacteria. Of 285 AI-generated designs, researchers synthesised and tested nearly 300 candidate genomes in the laboratory, resulting in 16 functioning phages that successfully infected and killed E. coli.
The AI-designed bacteriophages represent the first functional genomes composed by machine that have not been found in nature. Some of the designed phages were able to overcome bacterial resistance that defeated the original virus.
Reported impact
Experts noted that AI has entered the design stage of biology, which raises biosecurity concerns. One AI researcher described a potential threat of 'deepfake viruses'—viruses re-engineered by AI to maintain dangerous functions while rearranging genetic sequences to evade existing detection and screening systems.
Index analysis
The Stanford experiment demonstrates that computers are moving from analysing biological information towards proposing biological designs that scientists can physically build. While the creation of viral genomes and deliberate modification of viruses is not new to science, the use of AI to autonomously design functional novel viral genomes at scale represents a new capability in the field.
Governance controls
Experts noted that the development of AI-guided viral genome design requires appropriate governance and safeguards.
Classification
Sources and evidence
This record was researched and written by the Index, based on the following original reporting.