“Implies current safety controls are insufficient by suggesting mandatory intervention is necessary for AI systems.”