AI CEOs Warn Congress: Generative Models Are Lowering the Bar for Bioweapons

What They’re Actually Asking For

The letter, organized by the Foundation for American Innovation and the nonpartisan Institute for Progress, has a specific and concrete ask: mandate biosecurity screening for companies that sell synthetic DNA and RNA.

Right now, some sellers screen voluntarily. The letter wants that to become law — with a paper trail. Specifically, it calls for mandatory record-keeping on orders and material specifications, creating an audit layer that could support future biosecurity investigations.

Notably, some of the companies being asked to comply — including Twist Bioscience and Ansa Biotechnologies — also signed the letter. That’s not nothing.

Why Now

Generative AI reached 53% of the global population in just three years, according to a 2026 Stanford study. That’s faster than the PC. Faster than the internet.

Speed of adoption matters here because publicly available AI models have already demonstrated the ability to surface information on designing and spreading biological agents. The letter puts it plainly:

“the knowledge barriers which have historically prevented bad actors from obtaining biological weapons will meaningfully erode.”

That’s not a hypothetical. It’s a trajectory.

The Threat Isn’t New — The Leverage Is

Biological weapons have accounted for just 0.02% of historical terrorist attacks. But that low frequency masks a brutal risk profile. Anthrax, for example, carries a near-100% mortality rate when inhaled without treatment. The 2001 anthrax letter attacks — five deaths, 22 infected, one of the largest FBI investigations in history — came from a single individual with insider access.

AI doesn’t require insider access. That’s the shift.

Legal frameworks like the Biological Weapons Anti-Terrorism Act of 1989 and the post-9/11 PATRIOT Act expansions exist, but they were built for a world where synthesizing dangerous biological material required rare expertise and expensive equipment. Generative AI is changing that cost curve fast.

The Bill Already in Motion

Congress isn’t starting from zero. In February 2026, Senators Tom Cotton (R-Ark.) and Amy Klobuchar (D-Minn.) introduced the Biosecurity Modernization and Innovation Act of 2026, which would require synthetic material sellers to screen both orders and customers — with carve-outs for clearly non-hazardous materials.

The open letter is essentially a pressure campaign to accelerate that bill. Josh Wentzel, a senior fellow at the Foundation for American Innovation, called it “bipartisan, concrete, achievable, and noncontroversial” — which in Washington is practically a miracle.

The alignment between AI labs and the synthetic biology industry gives lawmakers a rare gift: a unified constituency asking for the same thing.

What This Means for the AI Tools Ecosystem

This isn’t just a policy story. It signals where AI governance is heading — and it has teeth.

If Congress moves on biosecurity screening, expect it to set a precedent for dual-use AI regulation more broadly. Tools that touch sensitive domains — biology, chemistry, cybersecurity — will face increasing pressure to demonstrate safeguards, not just disclaim liability in a terms-of-service footnote.

For founders building in or adjacent to these spaces, the message is clear: proactive compliance is becoming a competitive moat, not a burden.

Conclusion

The rarest thing in tech is competitors agreeing on anything. When they do — and when they’re asking for more regulation of their own products — it’s worth taking seriously.

The knowledge barrier is lowering. The question is whether the legal barrier rises fast enough to meet it.