Just six months after releasing its national AI framework, Tokyo is updating its guidelines to address the weaponisation of frontier AI models capable of finding and exploiting unknown vulnerabilities ...