It's likely a self solving problem. Through AI we are adding tons of fake information to the net. We will get to the point that a picture or video alone won't be definitive proof of anything. There will have to be back up evidence that it is real. I think a way to protect yourself from deepfakes is to go the opposite route and deepfake yourself. Make sure there are images out there that are provably fake, that way if one does come around that is not of your own making, you can point to a history of being deepfaked and be able to easily dismiss it.
Actually that's one of the arguments of the article. Seems reasonable that increasing awareness of Deepfakes in general will reduce the dangers.
An article about keeping yourself safe from Deepfakes from a seasoned Deepfaker.