[andrewflnr]

I doubt it will be that easy to bypass. A fake would still have to withstand pixel-level analysis on the level of methods that already detect tampering in regular video. For one thing, that will have to be a very high quality monitor indeed to leave no detectable trace of e.g. moire patterns.

[latexr]

A fake doesn’t need to be perfect to be effective, it just needs to fool enough people. Most posts won’t (and couldn’t ever) be scrutinised at that level.

[andrewflnr]

Even if a lot of people just lap up whatever tiktok feeds them, it still matters that people who care can get actual evidence from the real world that eventually filters out to public consciousness. It will have indirect influence, yes, but it'll still be a lot better than being fully post-truth.

[array_key_first]

Presumably if the video is AI generated we can also AI detect if it's a legitimate video.

That won't work for the original source probably, but I would think AI sophisticated enough to tell if you're just recording a monitor would exist.