Artificial intelligence: ‘One screw-up ... and we will destroy any remaining trust the public has in science’

Executives and biosecurity experts concerned that future models could help users create novel viruses or bioweapons