AI firms must be punished for ‘dangerous products’, says former regulator

Existing laws could see Anthropic, OpenAI and others face punishment for rogue systems