Alright I agree. Sounds like a realistic use.

Still feels like a gimmick. Particularly because most of the value you get out of processing the audio, not the video. And video is the part of deepfakes that I feel never should have been developed (but was inevitable, someone was eventually going to make it happen).

I'm not sure censorship is a better use of the tech. In terms of hollywood films I imagine it might be used to help make stunt/body doubles less obvious though