AI leaders call for tougher protections against AI-aided bioweapons

Leaders from rival AI firms, including Anthropic’s Dario Amodei, OpenAI’s Sam Altman, and Microsoft’s Mustafa Suleyman, have united in an open letter to U.S. lawmakers. They call for stricter regulations requiring companies selling synthetic DNA and RNA to screen orders for sequences capable of creating dangerous pathogens. The letter emphasizes that AI tools could make it easier for individuals to design and assemble harmful genetic material, lowering the expertise threshold for bioweapon development. The initiative also includes Meta’s AI chief Alexandr Wang, Google DeepMind’s Demis Hassabis, and executives from biotech firms like Twist Bioscience and Ansa Biotechnologies. The letter was reportedly organized by the Foundation for American Innovation and the Institute for Progress, reflecting broad concern across tech, science, and national security sectors. While some providers already screen orders voluntarily, the letter argues for mandatory oversight to prevent misuse and track potential threats. Scientists have long warned about the risks of synthetic biology enabling dangerous pathogen engineering or reviving extinct viruses. However, the concern now is that AI advancements could accelerate these capabilities, making them more accessible to those with limited resources. The letter states that current voluntary screening measures are insufficient and urges Congress to act swiftly to address the growing biosecurity gap. The signatories stress the urgency of the issue, noting that technological progress outpaces existing safeguards. They highlight the rare consensus among stakeholders who often disagree, urging policymakers to respond with decisive action. The letter also notes that AI could facilitate other threats, such as chemical weapons, broadening the scope of potential risks. While some synthetic DNA and RNA providers already maintain records of orders, the letter calls for standardized tracking systems to monitor any threats that bypass initial screening. The goal is to close loopholes that could allow dangerous sequences to be ordered and assembled without detection, mitigating the risk of accidental or intentional misuse.