OpenAI, Anthropic CEOs push Congress to regulate AI-bio weapons

The CEOs of OpenAI, Anthropic, Google DeepMind, and Microsoft AI have jointly signed a public letter urging Congress to pass laws requiring all companies that sell synthetic DNA and RNA to screen customers and orders. The letter, organized by the nonpartisan Institute for Progress and right-leaning Foundation for American Innovation, highlights that “there is a real possibility that the knowledge barriers which have historically prevented bad actors from obtaining biological weapons will meaningfully erode” as AI advances. It backs a bipartisan Senate bill that would mandate such screening for any gene synthesis provider operating in the United States. Currently, many providers rely on voluntary screening through the International Gene Synthesis Consortium, but not all vet orders or customers, and AI tools can help bad actors identify providers who don't screen or suggest modifications to evade detection. The signatories include scientists, national security experts, and executives from gene synthesis firms like Twist Bioscience.

However, screening tools have flaws. Last year, Microsoft researchers showed that AI protein design tools could generate dangerous gene sequences that slipped past existing screening software. David Relman, a microbiologist and biosecurity expert at Stanford, notes that AI can guide users to manipulate orders to bypass detection. Geoff Ralston, former Y Combinator president, argues AI labs with biology models should also screen users. The letter aims to close these gaps before AI-driven bioattacks become feasible. While bioterror incidents are rare, the potential for mass casualties and economic loss is immense—especially as gene synthesis costs drop and AI design capabilities improve. The proposed legislation would make screening mandatory, not voluntary, and would be a critical step in preventing an AI-enabled pandemic.