AI Designs 16 New Viruses, Sparking Bioweapon Concerns

Scientists have utilized artificial intelligence to design 16 novel, fully functional viruses, a development that has prompted concerns about the potential for creating dangerous bioweapons. The research, conducted at Stanford University, involved using AI to generate genomes for bacteriophages—viruses that infect bacteria.

Thousands of potential viral genomes were suggested by the AI, and researchers synthesized 302 in the laboratory. When tested against E. coli bacteria, 16 of the AI-designed viruses proved effective in killing the bacteria. While the study focused on viruses that target bacteria and poses no direct threat to human health, it demonstrates AI's capability to design viable, non-natural viruses. This breakthrough has been described as a significant turning point, with potential applications in disease treatment.

However, experts warn of the broader ethical implications and the potential for misuse. Dr. Simon Clarke, Associate Professor in Cellular Microbiology at the University of Reading, noted that while current safeguards within the AI model and the nature of the viruses used mitigate immediate risks, the underlying capability could be exploited. "This underlying capability means the technology could potentially be used to create dangerous pathogens with potential to be a bioweapon," Dr. Clarke stated.

He further clarified that AI can design genetic code but cannot physically create or release a virus independently. Sparking a pandemic would require human intervention, specifically the synthesis and assembly of a virus in a laboratory setting by malicious actors. "AI can only design genetic code, it cannot physically create or release a virus on its own," he told the Daily Mail.

Dr. Carl Mayers, a Senior Adviser in Biosecurity Policy at The Centre for Long-Term Resilience, echoed concerns about potential misuse but emphasized the current difficulty, expense, and time involved. "At the moment it would be extremely hard, slow and expensive to use this approach to make a bioweapon," Dr. Mayers explained. He pointed out that the researchers had to test hundreds of designs to find a small number of functional viruses against a simple bacterium, a process that would be considerably more challenging and costly for animal or human viruses.

Dr. Mayers reassured that an AI cannot initiate a pandemic on its own. The primary concern, he suggested, lies with controlling access to such powerful biological AI tools. "We should worry more about who has access to these powerful biological AI tools," he said. In beneficial applications, these tools could lead to new drugs and methods to combat antibiotic-resistant bacteria. Conversely, in the wrong hands, they could cause harm. Therefore, establishing governance frameworks and implementing robust safeguards is crucial, akin to those for other scientific and medical technologies, especially as laboratory automation advances.