AI leaders unite behind biosecurity: open letter urges mandatory nucleic acid synthesis screening
Key Points
- AI leaders including Sam Altman, Demis Hassabis, and Dario Amodei signed an open letter urging the U.S. government to mandate screening of nucleic acid synthesis orders for dangerous sequences.
- The International Gene Synthesis Consortium voluntarily screens roughly 80% of global synthesis capacity, leaving a 20% gap where orders can proceed without safeguards or government verification.
- AI tools could help bad actors generate pathogen sequences that synthesis companies might then assemble into working viruses, making mandatory screening a near-term supply-chain control rather than speculative risk management.
Summary
AI Leaders Back Mandatory Screening for Dangerous Gene Synthesis
A coalition of AI executives—including Demis Hassabis, Sam Altman, and Dario Amodei—has signed an open letter calling on the U.S. government to mandate that nucleic acid synthesis companies screen orders for dangerous sequences and keep customer records.
The threat is concrete. In 2002, researchers synthesized infectious poliovirus from published sequence data alone, without needing a physical sample. A 1918 Spanish flu virus was reconstructed the same way. The barrier to creating dangerous pathogens has shifted from needing biological samples to having the genetic blueprint—essentially text files containing ATGU sequences (the building blocks of RNA). With equipment increasingly available, that becomes a plausible risk.
AI amplifies it. The letter frames the urgency as coming from advances in AI. Without safeguards, AI tools could theoretically help generate sequences for dangerous pathogens, which bad actors could then order from synthesis companies and assemble into working viruses.
The current system has gaps. The International Gene Synthesis Consortium, established in 2009, says roughly 80% of global commercial nucleic acid synthesis capacity voluntarily screens orders. But membership is voluntary, reporting is self-reported, and the U.S. Department of Health and Human Services offers only non-binding guidance. Government agencies don't independently verify the consortium's 80% figure. The remaining 20% of synthesis capacity operates outside the consortium's safeguards.
The letter asks for three concrete changes: mandatory screening of orders for dangerous sequences, verification of customer legitimacy, and recordkeeping. These are standard supply-chain controls, not a new regulatory regime.
The signatories span AI labs (OpenAI, Anthropic, DeepMind, Microsoft, Y Combinator), nucleic acid synthesis companies (Twist Bioscience, ANSA), and biotech firms including Valthos. The letter is substantive—not another doomsday warning dressed as a press release, which has become a familiar pattern in AI announcements.
It's an early step. The open letter is a push for government action, not a binding commitment. Whether legislators act quickly depends on whether they treat this as a plausible near-term risk or a longer-horizon concern.
Every deal, every interview. 5 minutes.
TBPN Digest delivers summaries of the latest fundraises, interviews and tech news from TBPN, every weekday.