It is infuriating to see the crap posted as supposedly "real."

There are no reliable tools to detect use of AI.

I think a simple rule should be to just disclaim if AI is used. Posters can ignore it as they wish.

If the source is some obscure purveyor of videos seeking likes and clicks, and it contains material that looks either too slick or too cartoonish, assume it's AI. There are videos on YouTube and elsewhere that purport to be by Rachel Maddow and other well-known cable tv hosts that seem fairly real - using their voices and images, though sometimes the voices are too fast - but a real Maddow video will have the MS NOW logo, and these don't. If the video doesn't come from a known source and it looks slick or artificial it's AI. We are seeing a lot of elaborate cartoon-like "artwork" that is obviously AI. If something purports to be an actual photograph, it's usually too shiny; also, look for weird shadows and shapes, extra fingers or impossible overlays of objects. Be discerning. Don't post something just because you think it's clever. AI steals the images and the intellectual property of real people and churns it into crap.