For the first time, a team of researchers led by Stanford University has used generative artificial intelligence to design fully synthetic viruses unlike anything seen in nature.
This was done using genomic linguistic models trained on genetic data from bacteriophages that infect only bacteria and do not pose a threat to humans.
The study, led by Dr. Ursula Distrano, was published in the journal Nature Biotechnology on Wednesday and revealed the possibility of designing functional viral genomes using artificial intelligence, in a "crucial first step" towards using this technology in therapeutic applications.
The researchers used Evo1 and Evo2 models trained on two million phage genomes, deliberately excluding data on viruses that infect humans or animals to minimize risk. After generating thousands of potential genomes, they selected 302 designs for in vitro synthesis, 16 of which proved successful in killing E. coli bacteria, overcoming the resistance the bacteria had developed against natural phages. One of the engineered viruses, Evo-Φ69, demonstrated an exceptional ability to replicate, outperforming its natural counterpart in the competition.
Researchers believe this technology could open the door to developing new treatments to combat antibiotic resistance, as artificial intelligence could design phage viruses that target specific drug-resistant bacterial strains, paving the way for more effective personalized medicine.
However, this advance has raised deep security concerns, with biosecurity experts warning that the same technology could be used to design deadly human pathogens that current defense systems cannot detect.
Experts Thomas Inglesby and Moritz Hanke of Johns Hopkins University concluded that "the ability to synthesize viral genomes using generative artificial intelligence now exists; however, the governance necessary to guide them safely is lacking." Their research paper stated that AI could help "accelerate the design and release of highly virulent viruses."
While some experts believe the risk of misuse of this technology is exaggerated compared to the ease of modifying existing pathogens, others point out that the speed at which these technologies are developing outstrips the ability of governments to put in place adequate regulatory frameworks.
Dr. Distrano calls for "immediate governance" and "community oversight" of research before the technology gets out of control, at a time when previous publications about genetically modified avian viruses have attracted intelligence attention.
Researchers are calling for the development of proactive regulatory policies that adapt to the pace of technological progress, while engaging governments, scientists, and the private sector in global conversations about the responsible use of control over this new power.
Track
ReplyDelete