Published October 4, 2026
How diVine's AI Ban Actually Works
diVine's boldest promise is also its strangest: no AI-generated content, ever. In 2026, when every feed is flooded with synthetic video, that's a radical stance. But how do you actually enforce it? Not with moderation vibes — with cryptography.
The in-app camera rule
The core enforcement is simple: you can't just upload any video file. New clips must be recorded inside diVine's own in-app camera, which cryptographically signs each clip at the moment of capture using Guardian Project's ProofMode software. That signature is a tamper-evident seal proving the footage came from a real camera, at a real moment.
C2PA: the digital paper trail
For files that do come from outside, diVine checks C2PA provenance data — the content-authenticity standard backed by Sony, Canon, Adobe, Google, Meta, and OpenAI. C2PA metadata records whether a file originated from a camera or an AI tool. If the trail says AI-made, the upload gets blocked. It's a verification pipeline, not an after-the-fact moderation policy.
Why this matters
Most platforms fight AI slop with detection algorithms — an arms race they lose a little more each month. diVine flipped the problem: instead of detecting fakes, it proves realness. A signed, provenance-tracked clip doesn't need to pass a detector. As Rabble (founder Evan Henshaw-Plath) puts it, the six-second constraint plus the human-only rule makes the format "like a haiku" — precision over volume.
The limits
Honest caveats: no verification system is unbreakable, and the in-app-camera requirement adds friction that TikTok-style upload-anything apps don't have. Whether users accept that tradeoff at scale is diVine's biggest open question. But as a statement of values — human-made as a product category — it's the clearest line any social app has drawn.
New to the app? Start with our diVine guide or see how diVine differs from the original Vine.