Science From the Fringe • December 31, 2025 • Solo Episode
No guests identified for this episode.
In Episode 5 of In Defense of Virology, Bryce Nickels and Simon Wain-Hobson have a sobering conversation about the emerging risks of applying artificial intelligence to virus design.
The discussion centers on a recent Stanford preprint in which researchers trained AI models on bacteriophage genomes and used those models to generate entirely new viral sequences. From hundreds of AI-generated designs, the team synthesized 16 fully functional viruses. One replicated faster than its natural reference phage, while six displayed striking genetic stability, accumulating no detectable mutations at all. Even for Simon, whose career spans decades of studying viral evolution, the results were genuinely surprising.
Although bacteriophages are often framed as promising therapeutic tools, particularly for treating antibiotic-resistant infections, Simon cautions that the implications of this work extend far beyond phage biology. Applying similar AI-driven approaches to animal or human viruses could unintentionally generate pathogens that are more transmissible or more virulent—effectively producing gain-of-function outcomes without any traditional laboratory manipulation.
The episode places these findings within a broader and increasingly urgent context. Leaders in the AI community have begun sounding alarms about AI-enabled biology. Most notably, Yoshua Bengio warned in a New York Times op-ed that the implications of this technology are “terrifying.” Similar concerns were echoed at the September Red Lines AI meeting, where pandemic genesis was identified as the foremost global risk.
Simon and Bryce argue that this moment demands restraint rather than curiosity. They urge scientists to refrain from applying AI-based viral design to human-relevant pathogens and call on research funders, including the National Institutes of Health, philanthropies, and private foundations, to withhold support for work that could escalate existential risk. (Such restraint would align with recent U.S. executive orders aimed at preventing the enhancement of dangerous pathogens.)
The episode closes with Simon making a broader appeal for scientists to stop prioritizing technical novelty and high-risk experimentation, but for a renewed commitment to the principle of do no harm. Biology, he contends, should focus its energy on urgent human challenges (e.g., cancer, neurodegenerative disease, and mental health) where progress can be transformative without carrying the risk of catastrophic consequences.
(recorded December 7, 2025)