Even if there would be a simple way to define "lying" in this context and a simple 100% effective way to proof it. It would only shift the problem not solve it.

Even if there would be a simple way to define "lying" in this context and a simple 100% effective way to proof it. It would only shift the problem not solve it. You can already "lie" under oath if you formulate something as opinion if there is nothing that contradicts your statement, its that simple. If people can be sentenced for the writing words online if they intentionally lied that just puts a target on normal people an make professional writers team up with lawyers to avoid ever writing anything that could be deemed a lie. That solve no problem at all. People find a way to tell you that the earth is flat anyway. Putting wrong speaking closer to wrongdoings is a very dangerous idea in general. we should want more speak not less and we get that if speech is tolerated. The "inciting violence" thing is already very very close to breaking the concept of free speech. And it can also be defeated simply by linguistic tricks. "Kill the ...." would incite violence but "I think we should kill the ..." expresses an opinion. Also this very example here used the same words as something that in fact could incites violence but clearly my post isn't. Now do we really want an AI to detect de difference? Or maybe real human? Moderators who are almost certainly not qualified to judge because a content moderator isn't a judge and should not be.