Scientists Use AI to Design Viruses, Raising Legal and Bioethical Concerns
AI has been used to create synthetic viruses capable of infecting bacteria for the first time.
Why it matters: These developments expose gaps in existing biosecurity and patent laws, necessitating new regulatory and legal approaches. Corporate legal teams and biotech firms must navigate emerging liabilities linked to AI-driven biological innovation.
- A Stanford team created Evo-Φ2147, the first AI-designed synthetic virus targeting E. coli bacteria.
- FDA's CBER published draft guidance in January 2025 on AI use in biologics regulation.
- Health security experts warn current biosecurity oversight frameworks are insufficient for AI-driven biotech.
- Over 100 researchers at the Asilomar Conference called for strict data controls to prevent misuse such as bioweapons development.
A Stanford-led research group used generative artificial intelligence to design Evo-Φ2147, a synthetic virus capable of infecting and killing E. coli bacteria. This marks the first documented instance of AI creating a novel organism, raising profound bioethical and security concerns.
The development relied on the AI tool Evo2 and resulted in 16 distinct new viruses according to reports in Wired. Such rapid innovation has outpaced current regulatory mechanisms, prompting experts to call for urgent legal reforms.
Health security authorities, including Thomas Inglesby and Moritz Hanke from Johns Hopkins, emphasize that "the current frameworks for overseeing such high-risk bioscience are insufficient," highlighting risks of dual-use and potential misuse. Correspondingly, at the 50th anniversary Asilomar Conference, over 100 researchers endorsed stringent data controls to prevent AI applications that could aid development of bioweapons.
On the regulatory front, the FDA's Center for Biologics Evaluation and Research (CBER) published draft guidance in January 2025 addressing the safe use of AI and machine learning during the biological product lifecycle, signaling an institutional readiness to adapt oversight to this new reality (FDA CBER guidance).
With AI redefining capabilities in synthetic biology, industry and legal experts now grapple with questions around biosecurity liability, intellectual property rights, and ethical responsibility. As Princeton's Michael Hecht put it, "Machines are rethinking what it is to be human, what it is to be alive," underlining the weight of these emerging debates.
By the numbers:
- 16 new viruses created using AI — milestone in synthetic biology
- January 7, 2025 — FDA CBER draft guidance on AI in biologics regulation published
- 100+ researchers — endorsed data controls to prevent AI misuse at Asilomar Conference
Yes, but: Details on biosafety measures and environmental impact for AI-designed viruses remain limited, complicating risk assessments.
What's next: Stakeholders await formal adoption of FDA guidance and potential updates to international biosecurity treaties to address AI biotechnology.