Senior official describes AI Safety Institute approach to ‘bioweapons,’ plans for global network

A top official at NIST’s Artificial Intelligence Safety Institute described efforts to develop advanced testing methods for large language models, with an eye toward preventing the use of AI tools in making biological weapons, as well as plans for the fall launch of an international network of similar institutes, in a panel discussion hosted by the Center for a New American Security.

“We’re focused on how to ensure AI doesn’t make it easier to make bioweapons,” according to Stephanie Guerra,...