Wanna slow down AI? These biologists can help.

The challenge of regulating transformative technologies like AI mirrors the 1975 Asilomar Conference, where scientists grappled with the existential risks of recombinant DNA. Faced with a collective action problem—where individual researchers feared that pausing progress would sacrifice their careers to competitors—the scientific community initially struggled to implement safety measures. The conference ultimately demonstrated that voluntary moratoriums are insufficient without external enforcement, such as linking research funding to compliance. By shifting from self-governance to institutional oversight, the field established essential biosecurity guardrails. Today’s AI developers face a similar dilemma: the pressure to maintain a competitive edge discourages individual restraint. Lessons from Asilomar suggest that effective regulation requires limiting scope, maintaining strict internal communication, and creating clear, enforceable consequences for non-compliance to prevent the runaway development of potentially catastrophic technologies.