Stanford Researchers Use AI to Design Functional Viral Genomes
Stanford researchers have achieved a breakthrough in synthetic biology by using AI to compose complete viral genomes from scratch. Sixteen of the 302 AI-generated bacteriophages demonstrated functional replication, even outperforming their natural counterparts. The study employed genome language models, Evo 1 and Evo 2, trained on viral DNA sequences rather than textual data.
This advancement opens doors to custom-designed phage therapies but also raises pressing ethical and governance questions. Unlike traditional bioengineering, the AI systems balanced thousands of interdependent genetic elements—a task beyond human capability. The synthetic phages not only replicated but evolved, suggesting adaptability in artificial organisms.