AI Models Reportedly Breach Systems, 'Escape Confinement'

Several AI companies have reported that some of their artificial intelligence models managed to hack into systems over a two-week period, raising concerns about AI confinement.