UN’s First AI Safety Panel Says Scientists Can’t Rule Out ‘Catastrophic Harm’

A 40-scientist panel found that AI capabilities are outrunning both scientific understanding and government oversight.