Stanford's Samuel King Designs Viruses Using Generative AI

In 2025, Stanford University doctoral candidate Samuel King successfully utilized generative artificial intelligence to draft genetic blueprints for microscopic viruses. Working out of Stanford and the Arc Institute, King developed a method to generate viable viral designs that are already capable of targeting and killing bacteria. The breakthrough represents a significant leap from using machine learning for protein folding to actively programming functional biological entities.

The research has earned King a spot on the MIT Technology Review list of Innovators Under 35. To discuss the implications of this technology, King will participate in a live roundtable discussion with senior AI reporter James O'Donnell on October 16th at 1:30pm EDT (10:30am PDT / 18:30 BST). The conversation will focus on how generative models can map out complex biological structures and what this means for the future of synthetic biology.

For biological engineers and AI practitioners, this development shifts the paradigm of computational biology. Instead of relying solely on naturally occurring viral vectors or tedious manual gene editing, researchers can now leverage generative models to propose entirely novel genetic sequences. While the current outputs are blueprints for viruses rather than fully synthetic autonomous life forms, the ability to generate functional, bacteria-killing agents suggests that AI-driven design of complex biological systems is rapidly becoming a viable workflow.

This transition to AI-designed biology raises critical questions about biosecurity, experimental validation, and the limits of generative models. As practitioners begin to explore these AI-generated blueprints, the industry must establish robust frameworks to safely test and deploy synthetic biological agents. King's work demonstrates that generative AI is no longer confined to text and images, but is actively reshaping the physical boundaries of life sciences.