Prof Tom Inglesby advocated for formal independent evaluations of new artificial intelligence models prior to release to ensure biological dangers are eliminated.