AI Content & Provenance Statement

RapidThumbnails generates images and clips using machine-learning models. This Statement explains how we disclose AI generation, how content authenticity (provenance) is embedded in our outputs, what your obligations are when sharing AI content, and how we comply with applicable law.

1. Disclosure That Content Is AI-Generated

Every image and clip produced by the Service is AI-generated, in whole or in significant part. Where required by law, we apply a visible label to the output before download and embed cryptographic provenance metadata that downstream platforms (Instagram, YouTube, TikTok, and others adopting the C2PA standard) can read and surface to viewers.

2. C2PA Content Credentials

Outputs include Coalition for Content Provenance and Authenticity (C2PA) metadata, commonly known as Content Credentials. This metadata records:

• That the content was AI-generated.
• The AI model and provider that produced it.
• The date and time of generation.
• A cryptographic signature from RapidThumbnails so platforms can verify the metadata was not forged.

Stripping or tampering with this metadata before re-upload is prohibited by our Acceptable Use Policy.

3. Jurisdiction-Aware Labeling

Our labeling engine applies disclosure based on the jurisdiction you select in your account settings:

• European Union (EU AI Act, Article 50): visible AI label on the output and embedded provenance metadata.
• California (AB 2655, SB 942): visible label on AI-generated political content and Content Credentials on covered outputs.
• Other US states with AI-disclosure laws (current and emerging): visible label per applicable scope.
• Other jurisdictions: embedded C2PA only by default; visible label can be turned on per-generation.

You are responsible for selecting the jurisdiction that matches where your audience is and where you publish the content. The Service applies the most protective rule on ambiguity.

4. Your Obligations When Sharing AI Content

• When uploading to a third-party platform with its own AI-disclosure feature (YouTube, TikTok, Instagram), enable that platform's disclosure in addition to ours.
• Do not strip, edit, or replace the C2PA metadata before re-upload.
• Do not present AI-generated content as a real photograph or recording of an actual event without clear disclosure to viewers.
• Comply with platform-specific rules on synthetic media (manipulated-media policies, political-content rules).

5. What We Don't Promise

Our AI outputs are produced by third-party models. We do not warrant that:

• Outputs are free of artifacts, biases, or unintended likenesses.
• Outputs do not resemble existing copyrighted works.
• The C2PA metadata will be preserved by every platform you upload to (some platforms still strip metadata on upload).
• Visible labels are sufficient for every jurisdiction's requirements — we apply the standard most commonly accepted, but jurisdictional law is changing rapidly and may exceed our default disclosure.

6. Reporting Issues

If you encounter an AI output that:

• Depicts a real person in a way you believe lacks consent — submit a takedown via DMCA Policy (for copyright) or Acceptable Use Policy reporting (for likeness misuse).
• Reproduces a copyrighted work — submit a DMCA notice.
• Lacks correct labeling for your jurisdiction — email legal@rapidthumbnails.com.

7. EU AI Act Compliance

RapidThumbnails classifies as a deployer of a general-purpose AI system under the EU AI Act. We apply the obligations of Article 50 (transparency for AI-generated content) to all outputs and maintain the technical documentation required by the Act for our risk category. Our European representative will be designated as required.

8. Updates

AI disclosure law is evolving rapidly. We will update this Statement and our labeling engine as new rules take effect.