AI leaders urge US Congress to regulate synthetic DNA to prevent bioweapons

Leading artificial intelligence executives are calling on US lawmakers to move quickly on new safeguards that would make it harder to use AI and synthetic biology to develop biological weapons. In a rare joint intervention, heads of some of the most influential AI labs want Congress to mandate screening of orders for synthetic DNA and RNA.

The appeal comes amid growing concern that increasingly capable AI systems could help hostile actors bypass traditional knowledge and technical barriers in the field of biology, unless stricter controls are introduced.

AI companies warn of eroding barriers to biological weapons

In an open letter to Congress, senior figures from OpenAI, Anthropic, Google DeepMind, as well as representatives of Meta and Microsoft, argue that rapid AI progress is changing the risk landscape for biological threats.

Among the signatories are Anthropic CEO Dario Amodei, OpenAI CEO Sam Altman and Google DeepMind chief Demis Hassabis. They caution that AI systems are advancing so quickly that the knowledge barriers which historically made it difficult for malicious actors to acquire biological weapons “may be significantly weakened”.

Call for mandatory screening of synthetic DNA and RNA

The letter urges Congress to pass legislation that would require companies producing and selling synthetic nucleic acids – DNA and RNA – to carry out systematic checks on all orders.

According to the proposal, suppliers should:

• screen ordered sequences for potentially dangerous genetic combinations that could be used to create biological agents, and
• verify the identity and credibility of customers placing such orders.

Currently, sequence screening is largely voluntary and practiced only by parts of the industry or by entities that receive US federal funding. Advocates of new rules argue that, given the capabilities of modern AI tools, relying on voluntary standards is no longer sufficient.

Regulation debate in Washington intensifies

The appeal coincides with a new executive order signed on Tuesday by President Donald Trump, focusing on oversight of AI models and cybersecurity. As reported by the Wall Street Journal, this marks a shift from the administration’s earlier, strongly deregulatory approach that emphasized winning the AI race against China.

The final version of the executive order was reportedly softened in terms of requirements for AI safety testing, after pushback from large technology companies including Meta. Nevertheless, discussions about stricter oversight have continued.

Sam Altman presses for tougher AI requirements

Despite the diluted language of the executive order, OpenAI chief executive Sam Altman met on Wednesday with White House officials and members of Congress to present the company’s ideas for tighter obligations on developers of advanced AI models, the Wall Street Journal reports.

The newspaper notes that Trump previously revoked an order issued under President Joe Biden which had set out a framework for gene synthesis verification. The current White House has announced that it is working on its own guidance for this area, but has not yet published detailed rules, stating only that it seeks to “balance innovation and security”.

Supporters and critics of new law clash over scope

Proponents of a statutory solution argue that executive orders alone cannot provide comprehensive protection, because they typically do not cover all buyers and suppliers of synthetic DNA and RNA. Binding legislation, they say, would create uniform obligations across the entire market and help close potential loopholes that AI-enabled actors might exploit.

Opponents of new legal requirements counter that defining precisely which nucleic acid sequences should be classified as dangerous is inherently subjective and may evolve as science progresses. They also warn that compliance costs and administrative burdens could fall hardest on smaller, innovative companies in the synthetic biology sector, potentially slowing down beneficial research and commercial development.

The dispute in Washington illustrates a broader global challenge: how to design governance for powerful AI systems and biotechnologies that mitigates catastrophic risks without stifling innovation. The open letter from AI leaders signals that, at least in the area of synthetic DNA screening, parts of the tech industry are now openly asking lawmakers to move from voluntary norms to enforceable rules.