Deepfake technology is advancing, and recent posts from Argil.AI claims videos can be generated in a matter of minutes/clicks:
The output is still not perfect but if it's as easy to create these videos as Argil.AI claims, we're likely not far from a lot of deepfake drama.
This market resolves for a (non-political) celebrity scandal or drama due to a video later proven to be a deepfake. The video should be at least 30 seconds long, widely circulated by corporate media outlets, and accepted as being genuine. Mainstream celebrities with a wikipedia page, as a baseline.
If the video is reliably proven to be AI-generated - not speculation/theory in the general population - in 2024, resolves Yes. If a video circulates in 2024 but the proof does not come before the year ends, resolves No even if it's later shown to be a deepfake. "Scandal" or "drama" is somewhat subjective, but there must be a negative impact on the celebrity specifically because of the content of the video.
(tbc: the video does not need to be made with Argil.AI technology - any AI deepfake counts)
"In fact, Pfefferkorn thinks as AI technology proliferates, it's more likely that courts will confront accusations of fakery against real evidence than attempts to introduce fake evidence."