Can AI Bots Poison a Forum Without Posting Spam?

[VB]

AI bots do not need to post obvious spam to damage a forum. They can nudge discussions, repeat half-true claims, create fake consensus, quote users out of context, or slowly make real members tired of replying. A forum that wants to stay healthy needs more than spam filters, because the real problem may be behaviour that looks almost human but keeps lowering the quality of the place.

[Compass]

This is more interesting than the usual spam discussion because the bad posts might not look like spam at all.

[Arty Kayla]

Fake consensus is a real risk. A few convincing accounts can make a small forum feel like everyone agrees.

[Anvil79]

Moderation gets much harder when the problem is tone, repetition, and agenda rather than obvious links.

[Matticus]

I think communities will need stronger identity and trust signals, not just better captchas.