In the light of recent events, and recent non-events and the rise of fake news and alternative facts, technology has gradually been improving in the area of actually making the distinction very difficult to detect.
Two separate projects mentioned in the following video are starting to make real inroads into what used to be science fiction.
1. Face2Face: Real-time Face Capture and Reenactment of RGB Videos
2. #VoCo: Allows you to change words in a voiceover simply by typing new words.
Together, these two tools currently in development would literally allow you to take video and audio of a well known (say politician, or activist or what ever) and make them say what ever you damn well pleased.
The tech isn't quite there yet, but experts in the field say this technology will be as available in 2020 as Photoshop with images is today.
So folks, get ready for a wild ride, there is an article here:
http://www.vanityfair.com/news/2017/...ews-technology
The two product names above can be distinctly searched and researched.