Page 2 leans into the thing people always ask about: none of this is a deepfake. A deepfake maps a real face onto a real body, while these clips are generated whole cloth by diffusion models, so nobody real is depicted. Browse the set and judge the difference for yourself.
The deepfake question deserves a clear answer: a deepfake reuses a real human, while these clips invent the subject from nothing. That difference is why we can host the whole set in the open. Every figure on this page is a fabrication with no real counterpart.
Keep browsing: open the full AI video archive index, or dig into the Collision synthetic AI set and the Aftershock AI clip collection. The archive currently holds 1483 synthetic clips across 30 pages, every one AI generated with no real people.