Skip to content
All policies

AI disclosure

How every Clipsy video is marked as AI-generated and how anyone can verify it.

Last updated September 7, 2026

In short

Every finished video is marked in three ways: a machine-readable manifest inside the file, the AI flag each platform offers, and a public verification page. A visible badge in the video is added whenever a real person's face or voice is synthesized or when generated video makes up most of the reel.

The machine-readable manifest

Each video carries a deterministic JSON record, version clipsy-1, written into the MP4 metadata tree and into the file comment. It lists the models used per stage, whether a consented voice or face was used, whether a visible label is on, the platform flags, and the address of the verification page. Equal videos produce equal records. Clipsy does not sign this record with C2PA or add an imperceptible watermark in this release.

Platform flags

Every post asks the platform to set its own AI marker: TikTok's AI-generated flag, Instagram's AI label, and YouTube's synthetic media declaration. These are always on and cannot be turned off in settings.

The visible badge

A short badge reading "AI-generated" in the video's language is always on when a real person's face or voice is synthesized, when a project uses a person avatar, when generated video is more than sixty percent of a promo reel, or when a daily short is entirely generated. For other videos, an account setting chooses between always and when required. Accounts in EU or EEA timezones default to always.

The walkthrough angle

A walkthrough reel opens with an on-screen line in its first three seconds saying it is the owner's own AI demonstration. That line is recorded in the manifest as the disclosure overlay.

The verification page

Every finished video has a public page at /verify/ followed by its identifier. It shows the creation date, the models per stage, the three platform flags, whether a visible label is on, and a yes or no for likeness use. It shows no consent identifiers and no account data.

Disclosure responsibilities

Article 50 of the EU AI Act sets transparency duties for generated content and deepfakes, with applicability and transitional rules depending on the system and use. The FTC prohibits fake or false reviews and testimonials; it does not ban every use of an AI avatar. Clipsy prohibits synthetic consumer testimonials as a product rule. Review the requirements for your audience and platform; labels alone do not establish legal compliance.

Official sources

EU 2024/1689, Article 50FTC

Operator and contact

Clipsy is operated by Selmi Abderrahim in Tunisia.

For support, privacy requests, billing questions, or a report about content, contact:

support@clipsy.video