He could post an encrypted version and post the key with it to avoid it being trained on?

This sounds like paranoia to me to be honest. Please tell me I'm wrong.

I could have easily come up with just the same claim, without seeing the benchmark, it doesn't exist.

Maybe if we weren't anonymous and your profile leads to credentials that you have experience in this field, otherwise I don't believe it without seeing/testing myself.

but they've asked all the AI models this question. Whatever you tell an AI model is also in its training data

why?

you already sent the prompt to gemini api - and they likely recorded it. So in a way they can access it anyway. Posting here or not would not matter in that aspect.