In a not so distant future we might need to have some sort of regulation that forces uploaders (or content creators) to declare if videos have been generated with ai tech or not and depending on the content such declaration might carry legal consequences. On the other side hosting platforms should display clearly if such content was declared ai generated or not as well. Right now I can't see a simple and good enough solution as this that could mitigate the spread of malicious content.
My hunch is that it should be easy to decipher if it’s a AI video based on how the frames transitions.
Or we will have some cryptographic going on that is connected with your ID to prove that you are indeed human.
Trust on a society level is some other beast of difficult problem.