U.S. Scientists Are Using AI to Design Viruses, and people are not taking it well
Artificial intelligence is moving deeper into U.S. scientific research as labs race to improve vaccine development, outbreak forecasting, and drug discovery. In that push, researchers at U.S. universities and federally funded institutions are also using AI to help design viruses or viral components, a line of work that has drawn renewed public attention in 2024. The work is legal and often framed as biodefense or basic research, but it has also reopened questions about how far these tools should go and who oversees them.
Researchers are using AI tools to model and design viral systems
U.S. scientists have publicly described using AI to generate or refine viral proteins, predict mutations, and in some cases help design viral genomes, according to papers published in journals including Science, Nature Biotechnology, and Cell over the past two years. On January 18, 2024, the RAND Corporation published an analysis stating that generative AI could lower technical barriers in biological research, including work involving pathogens. That report did not say AI has already made dangerous virus creation easy, but it did say the capability trajectory deserves policy attention.
Federal agencies have also acknowledged the overlap between AI and biosecurity. The White House executive order on AI, signed on October 30, 2023, directed agencies to develop standards for screening nucleic acid synthesis and reducing biological risks tied to AI-enabled design tools. Researchers and institutions involved in this area have said the goal is to better understand viral behavior, improve countermeasures, and speed medical research, not to create public health threats.
The impact is national, but local details are still limited
This is not tied to one state or one single lab. Institutions in Massachusetts, California, Maryland, and Texas have all published virus-related AI or computational biology research in recent years, based on public university announcements and journal records. What is confirmed is that the research base is national and includes both academic and government-connected work, but a full public list of every U.S. lab using AI in virus design is not available.
That gap matters because oversight in the U.S. is split across agencies and institutions. The National Institutes of Health, the Department of Health and Human Services, and local institutional biosafety committees all play roles, depending on the project, according to federal guidance. What is not yet known is how many active projects involve generative AI specifically, how many reach the stage of physical synthesis, or how often proposals are rejected before lab work begins.
The bigger issue is oversight, not just the software
The current debate is happening because AI tools are improving quickly while U.S. biosecurity rules are still catching up. In March 2024, the National Security Commission on Emerging Biotechnology warned that AI could accelerate biological design and recommended stronger screening, clearer reporting rules, and updated federal coordination. The World Health Organization also released guidance in 2024 calling for governance frameworks around AI use in life sciences.
For the public, the practical takeaway is that most of this work remains inside regulated research settings, not consumer products or routine health care. There is no federal notice showing broad public exposure tied to AI-designed viruses, and agencies have not announced a nationwide emergency related to this research. What residents should expect next is more federal rulemaking, more disclosure requirements for DNA synthesis providers, and more debate over where legitimate research ends and unacceptable risk begins, based on ongoing U.S. policy reviews in 2024.
