Feature · Check
Disclosure in code
The AI disclosure is enforced in code, at the boundary where an ad leaves the system, not suggested where the caption is written.
Why it is a control and not a prompt
A synthetic presenter who reads as a satisfied customer can be treated as a fake testimonial under the FTC Endorsement Guides and the 2024 Rule on Consumer Reviews and Testimonials. A prompt is a request; a check is a control. So the script prompt forbids first-person experience claims, the output is checked afterwards, and the publish step refuses any payload whose caption lacks the disclosure.
What the caption says
Every ad on this site ships with it. The demo presenter ad carries: "Dramatization. Presenter is AI-generated, not a real customer." Replies to comments and messages pass the same first-person-claim gate the scripts do.