Is your chatbot exposed under Article 50?
Free scan for undisclosed AI chat and unlabeled synthetic content — the two failure modes national regulators are expected to enforce first, because they're the easiest to prove.
What this actually checks
Does the page load a known chat-widget script with no visible "you're talking to an AI" language anywhere nearby?
Does it reference AI media-generation tooling (Sora, Midjourney, Stable Diffusion, etc.) without a machine-readable content-credentials/C2PA marker?
Both checks read server-rendered HTML only — widgets injected client-side after page load can be missed. That's why every result says "decision support," not "certified." Full method, in one paragraph, no black box.
Get monitored, not just scanned
Continuous monitoring and the Enterprise tier (GPAI documentation, ISO 42001 audit prep) are launching to design partners first. Join the list — no payment collected here or anywhere on this site yet.
Who's behind this, and other questions
Who's behind this?
Pre-launch, solo-built, not yet an incorporated company — a Belgian entity is in progress. Said plainly here instead of hidden, because a compliance tool that isn't straight about its own status isn't worth trusting.
How do I reach you?
[email protected] — a dedicated address, not a personal inbox. It forwards privately; nobody's personal email is exposed here.
Is this legal advice?
No. It's a heuristic scan, not a legal opinion. See the Terms & Disclaimer for the full statement.
What happens to the URL I scan?
Logged for abuse prevention and to improve detection accuracy, deleted after 90 days. Full detail in the Privacy Policy.
Why is the free scan free?
It's the whole product's proof of work — if the free tier can't find a real, citable issue, there's no reason to trust the paid monitoring tier either.