Anthropic is very focused on AI safety. It makes LLMs safe by shutting down anyone from using them
It's the best way to ensure model wellness