The more interesting question remains unanswered, however. What would possibly be the reason they would find those particular posts potentially sensitive? It seems as though the visual processing NNs they apply to pictures sensed some kind of adversarial content in what are perfectly innocent pictures (one of them is a picture of my room).

I don't expect a reply, I just thought I'd throw this out there. It intrigues me.

Yes. I unchecked it.

Thank you very much for your help.

I'd say so. But I mean, it's probably because the account looks like a bot posting random images and tagging big accounts.