Do you think it was right for OpenAI to use AI to draft its official security-incident notification?

OpenAI's legal and security team used AI to help draft the email notifying Australian authorities about a breach of four systems in June. The company learned of the incident in August but only reported it on September 10; the message was reviewed and sent by humans.

• It's fine, it's just an efficient writing tool
• It bothers me that the company at fault writes its own excuse with AI
• No problem, as long as humans check the result
• The six-week delay is the bigger issue, not the AI