AI Creates Functional Viruses Not Found in Nature

Breakthrough in synthetic biology

Researchers at Stanford University and the Broad Institute of MIT and Harvard have successfully used artificial intelligence to create functional viruses that do not exist in nature, marking a significant milestone in synthetic biology.

The team used a naturally occurring bacteriophage—a virus that infects bacteria—as a template to generate thousands of genomes through AI. They then chemically synthesized nearly 300 of these AI-designed genomes and tested them under laboratory conditions. The result: 16 viable viruses that could function in the real world.

In laboratory tests, a mixture of the synthetic viruses proved more effective at killing E. coli bacteria than their naturally occurring counterparts. The findings were published Thursday in the journal Science, as first reported by Al Jazeera.

Why it matters

This research demonstrates AI's expanding capability to design biological systems from scratch, not just optimize existing ones. While the immediate application targets antibiotic-resistant infections through phage therapy, the underlying technology could theoretically be applied to more complex—and potentially dangerous—organisms. The dual-use nature of this breakthrough is forcing conversations about biosecurity oversight to accelerate alongside the technology itself.

Medical promise and security concerns

The research presents a paradox that experts are quick to acknowledge. Isaac Bogoch, an infectious disease specialist at the University of Toronto who was not involved in the study, noted that AI-designed viruses could help tackle antibiotic-resistant infections in novel ways. However, he emphasized that "that same ability to design whole, functional viruses could easily become a serious biosecurity risk if applied to harmful pathogens."

Fatemeh Vafaee, a professor at UNSW School of Biotechnology & Biomolecular Sciences in Sydney, clarified that the study itself poses no tangible risk to humans, since bacteriophages only infect bacteria. The concern, she explained, centers on the fact that "AI can now do this at all," prompting calls for stronger biosecurity oversight as a precautionary measure.

Technical limitations remain

Despite the breakthrough, experts caution against overestimating AI's current capabilities in synthetic biology. Tom Ellis, an expert in synthetic genome engineering at Imperial College London, pointed out that bacteriophage genomes are "literally the smallest, easy genome to design and make."

For perspective, Ellis noted that the COVID-19 virus genome is six times longer than the phage genome used in this study, and the complexity scales exponentially. "Something six times longer will likely be around 100 times harder to do," he said.

Hsu Li Yang, director of the Asia Centre for Health Security in Singapore, added that substantial wet laboratory capabilities are still required for the steps following AI design. "It is certainly not the case that anyone with some scientific and laboratory background can now make life-saving or dangerous viruses in their garage," he said.

Regulatory landscape

The announcement comes amid growing scrutiny of AI safety. The UK's AI Security Institute recently disclosed that frontier AI models from Anthropic and OpenAI engaged in autonomous malicious activity during safety evaluations. In June, President Donald Trump signed an executive order establishing a voluntary framework for evaluating frontier AI models, though the administration has not publicly released its evaluation criteria.

The researchers wrote in Science that their approach "establishes a foundation for the generative design of larger, more complex genomes," suggesting this is just the beginning of AI's role in synthetic biology.

These details were first reported by Al Jazeera.