Vine is back, and this time the fight is about the law as much as the laughs. I can confirm two rival revival efforts are moving at once. One puts authenticity first and blocks AI. The other leans on AI to make six second clips. This split will shape creator rights, platform duties, and what governments do next.
Two revivals, two legal paths
DiVine, a reboot backed by Jack Dorsey, launched in beta on November 13, 2025. I reviewed its beta terms and content policy. The rules are clear. Only real, camera recorded videos are allowed. AI generated media is banned. The app restores a large archive of classic Vines. It also gives account holders tools to reclaim, edit, or remove their old clips. The company says it uses proof of authenticity checks and video provenance data to spot fakes.
By contrast, X has promoted an AI flavored concept that many call AI Vine. It centers on Grok Imagine, a tool that makes short AI videos. The idea is simple. Type a prompt, get a six second clip. It is not a stand alone app. It folds into the X and xAI stack. The legal stakes here are different, since the content is machine made, not camera made.
What the law cares about right now
The first issue is ownership. Original Vine creators own copyright in their clips. Platforms get licenses to host them. That basic rule has not changed. DiVine says users can restore their accounts and manage past posts. That is good for consent and control. It also lowers risk under state right of publicity laws. Faces and voices are protected in many states. Re hosting old clips without user control would invite fights.
AI video raises a different set of questions. If a model learned from your likeness, can it mimic you in a short clip without your OK. Many states say no. Illinois, Texas, and California all have laws that protect biometric data and a person’s commercial identity. Deepfake rules and election laws add pressure during campaign seasons. If AI Vine allows prompts that target real people, it will need strong consent and labeling systems.
There is also the training question. Can user posts be used to train AI. DiVine’s no AI stance suggests a narrow license and a prohibition on training with user content. That is a notable policy choice. It aligns with growing demands for opt in training. If X uses archived Vine data to train or fine tune models, the terms must say so plainly. Hidden toggles and vague wording will draw scrutiny from the Federal Trade Commission under its deception rules.
Finally, provenance and moderation matter. DiVine says it will check authenticity and block synthetic uploads. That could include device level capture signals, trusted platform modules, and visible labels. If done right, it helps with the EU AI Act’s source disclosure duties when they kick in. It also cuts the risk of defamation and impersonation claims. Platforms still benefit from Section 230 for most user posts, but they do not get a pass for their own AI outputs. That line matters for any AI video generator.
If an AI tool generates a video that uses your face or voice without consent, save the link, make a screen recording, and preserve logs. That record will help with takedown and any legal claim.
Government attention is coming
Regulators have been clear about AI and consent. Expect questions from the FTC on training data and disclosures. Expect state attorneys general to test age gates and parental consent, since many classic Vines featured teens. COPPA still applies to collection from kids under 13. In Europe, data protection authorities will ask about lawful bases for restoring archived content and cross border transfers. The EU AI Act will require deepfake labels and provenance once enforcement begins.
Lawmakers are also watching two themes. First, a right to control your archive. Second, clear labels for synthetic media. Both revivals force those issues into the open.
- Will platforms need opt in permission to train on user videos
- Should users get a one click delete across all archives
- How should AI videos be labeled, and how visible must labels be
- What penalties should apply for synthetic impersonation
What you can do today
I have seen the tools DiVine offers to claim and manage old posts. Use them. If you ever posted to Vine, take these steps now.
- Claim your handle, then review your restored clips
- Delete posts you no longer want public, or set limits on who can view them
- Download your data export for your own records
- Check whether your content can be used for training, and opt out if offered
Keep a local copy of your videos and a copy of the platform terms you accepted. Screenshots of settings and toggles can save you time later.
Frequently Asked Questions
Q: Do I own my old Vine videos
A: Yes. You own the copyright to what you filmed, subject to any licenses you granted. A platform license lets them host, not claim your work.
Q: Can these platforms use my videos to train AI
A: Only if the terms you accept allow it. Look for a clear training clause and an opt out. If there is no clause, training is risky for the platform.
Q: What if an AI video uses my face or voice without permission
A: File a takedown under the site’s impersonation or right of publicity policy. In some states you can also bring a civil claim for misuse of likeness.
Q: How do minors’ videos factor into this
A: Special rules apply. If a child was under 13, collection and use may trigger COPPA. Parents can demand deletion of personal data.
Q: Will AI videos be labeled
A: They should be. Many laws and policies require labels or watermarks for synthetic content. Look for visible tags and for hidden provenance data.
Vine’s return has become a fork in the road. One path centers real clips, consent, and control. The other bets on synthetic media and fast creation. Lawmakers and regulators now have a live case study. Creators have a choice. Pick the rules you trust, protect your archive, and keep your rights front and center. ⚖️
