OpenAI, Anthropic push Congress to regulate synthetic DNA sales
In brief
- OpenAI, Anthropic, Google DeepMind, Microsoft urge Congress to regulate synthetic DNA and RNA sales
- Proposal mandates vendor screening against dangerous sequences and customer verification
- Mandatory compliance could consolidate market toward larger vendors, creating friction for decentralized science
The regulatory proposal
DNA and RNA vendors would need to screen every order against databases of known dangerous sequences, implement mandatory customer verification, and conduct comprehensive risk assessments before shipment. The open letter frames these requirements as necessary safeguards in an era when AI can assist in evaluating biological threats.
OpenAI has been conducting internal red-teaming exercises since early 2024, specifically testing whether its large language models could assist in navigating complex biological processes. Anthropic's own research has flagged the potential for large language models to provide meaningful uplift to individuals attempting to work through difficult biotech workflows. Both companies view this as a credible risk vector.
Market and innovation implications
Mandatory screening requirements carry uneven consequences. Smaller vendors with thin margins could find the new compliance costs burdensome enough to consolidate the market toward larger firms with existing infrastructure. Meanwhile, biosecurity technology companies building screening tools and sequence-analysis platforms could benefit directly from federal mandates creating demand for their products.
Mandatory screening requirements at the DNA synthesis layer could impose new friction on workflows that decentralized science projects depend on. DeSci communities emphasizing open-access biological research may face additional barriers to participation and collaboration.
Competitive rivals align on risk
The fact that Amodei and Altman, whose companies are locked in competitive rivalry, co-signed the same letter tells you something about how seriously they view the biosecurity risk. This initiative aligns with broader AI policy momentum observed in 2026, building on earlier executive actions that began establishing dual-use technology oversight frameworks.
The letter represents a rare convergence of major AI labs on a biosecurity issue—one that could reshape how federal policy approaches the intersection of AI capability and biological risk.
