Synthetic DNA screening is becoming a new focus for AI safety policy after major AI leaders backed rules aimed at reducing bioweapons risk. According to CNET, the heads of OpenAI, Anthropic, Google DeepMind, and other AI groups signed an open letter warning that better models could lower the barrier to biological misuse.
The letter is not asking Congress to regulate every AI model directly. Instead, it supports mandatory screening and recordkeeping for companies that sell synthetic DNA and RNA. That matters because gene synthesis firms can print custom genetic sequences used in medicine, diagnostics, and research, but the same supply chain could be abused to order dangerous biological material.
Synthetic DNA screening targets the supply chain
WIRED reports that signatories include Google DeepMind CEO Demis Hassabis, OpenAI CEO Sam Altman, Anthropic CEO Dario Amodei, and Microsoft AI CEO Mustafa Suleyman. The letter was organized by the Institute for Progress and the Foundation for American Innovation.
The risk is not hypothetical only in theory. WIRED notes that Canadian researchers used about $100,000 worth of mail-order DNA in 2017 to reconstruct the extinct horsepox virus, raising concerns that similar techniques could be used for more dangerous pathogens. Gene synthesis has also become cheaper since then.
The proposed rules would require providers to screen both customers and orders. Many reputable companies already check orders for sequences of concern, but the system is still partly voluntary and uneven. A bipartisan Senate bill introduced earlier this year would require U.S. gene synthesis providers to screen orders and customers.
There is also a technical gap. Microsoft researchers reportedly showed last year that AI protein-design tools could generate potentially dangerous sequences that slipped past some screening software. That is why AI companies may also need model-side safeguards, not just rules for DNA vendors.
The broader takeaway is clear: as AI makes biology easier to navigate, policy makers are being pushed to secure the physical bottlenecks before the risk moves from theory to practice.
You can follow more developments in Technowatt’s Artificial Intelligence coverage.