Australia flags criminal law changes to create 'failure to prevent' AI harm offence
The rapid review recommends Australia's Attorney-General's Department consider penalties for egregious regulatory non-compliance and a standalone offence criminalising AI models that cause harm without reasonable preventive steps.