Not only that they can review flagged content, but they tend to have separate retention policies for flagged content. Anthropic's is this: "We retain inputs and outputs for up to 2 years and trust and safety classification scores for up to 7 years if your chat or session is flagged by our automated trust and safety systems as violating our Usage Policy."

So don't run for office or anything like that. Someone, somewhere will have a contact that will get that.