AI model blocking criteria raises questions about safety and risk management in AI development.