Oops, true. I regard Anthropic and OpenAI as essentially the same company. Neither of them seem capable of securing their infrastructure.