Home › Store › Disclosure audit
Your chatbot has been breaking EU law since August 2.
On August 2, 2026, the EU AI Act's transparency rules went live. If your chatbot talks to visitors without telling them it's AI, or your site publishes AI-generated content with no marking, you're out of compliance — and the remaining grace period for content marking runs out December 2, 2026. We check your disclosures and write the fixes. Fast, fixed price, no lawyer invoice.
$149 audit. $349 with implementation.
The audit gives you the full disclosure checklist for your site: what's missing, exactly what to add, and where. The $349 tier adds us writing and installing every disclosure for you. Email-only, async, done in days.
Check my disclosuresWhat the law actually says
Article 50 of the EU AI Act is the transparency article. It doesn't care about your risk tier — it applies to what the system does. Four duties, plain English:
- Chatbots must say they're AI (Art. 50(1)). If your AI system talks directly to people, it has to tell them it's AI — unless that's obvious from context. A chat widget with a human name and no label fails this.
- AI-generated content must be marked (Art. 50(2)). Synthetic audio, image, video, or text has to be marked as artificially generated in a machine-readable format. Systems already on the market got a grace period for this marking — it ends December 2, 2026.
- Emotion-recognition and biometric systems need notice (Art. 50(3)). Anyone exposed to them must be informed first.
- Deepfakes must be disclosed (Art. 50(4)). AI-generated video of real people, and AI-written text published as news, has to be labeled.
Penalties under the Act's framework run up to €15 million or 3% of global annual turnover for the most serious breaches. (Full text of Article 50.)
The US isn't sitting out either
- Texas — TRAIGA, in force Jan 1, 2026. Applies to anyone doing business in Texas or serving Texas residents. Civil penalties $10,000–$200,000 per violation, enforced by the Attorney General (with a 60-day cure period). (LexBlog: TRAIGA now in effect)
- New York — synthetic-performer ad law (June 9, 2026). AI-generated likenesses in advertising need disclosure.
- Connecticut — LLM-training disclosure requirements landing in privacy notices.
- Colorado — AI Act, enforcement live June 30, 2026. Impact assessments and discrimination reporting for high-risk systems.
This is a patchwork and it's still moving. The audit covers the disclosure layer — the part every small business can actually fix this week.
What's in the $149 audit
- Chatbot disclosure check. We read your chat widget, voice setup, or AI assistant as a visitor would. Does it disclose it's AI at first interaction? Is the label clear and distinguishable, or buried in the terms? You get the exact wording to use.
- Synthetic-media marking check. AI-generated images, videos, or voice on your site or ads — we check what's marked, what's missing, and what machine-readable marking you need before December 2.
- Privacy-notice paragraph. A ready-to-paste paragraph disclosing AI use and LLM-training data practices for your privacy policy.
- 1-page risk memo. What applies to you, what doesn't, and what to fix first — in plain English, no legalese.
The $349 tier: we implement it
Everything in the audit, plus we write and install every disclosure for you: the chatbot label, the synthetic-content marks, the privacy-notice paragraph dropped into your policy. You approve the wording, we ship it. Done in 3 days.
What we need from you
- Your website URL.
- What AI tools you run customer-facing — chatbot, voice agent, AI image/video generation, anything.
- Link to your privacy policy (if you have one).
The terms, plain
- $149 flat for the audit, $349 flat with implementation. Paid up front.
- Audit delivered in 3 days. Implementation done in 3 days after you approve the wording.
- No calls. Everything runs async by email.
- Cash App or PayPal. Payment details when you claim your slot.
- This is a documentation and checklist service, not legal advice. We're not lawyers. If your situation needs legal counsel, talk to one — our memo will tell you when it does.
Questions
I don't have EU customers. Do I need this?
Maybe not the EU part — but the US state layer (Texas, Colorado, New York, Connecticut) applies to plenty of US-only businesses, and customer trust doesn't care about jurisdiction. If your bot pretends to be human, your customers will find out. The audit tells you what actually applies to you.
My chatbot comes from a vendor (Intercom, Zendesk, etc.). Isn't disclosure their problem?
Article 50 splits duties between providers and deployers. The vendor builds the system; you deploy it on your site. The disclosure to your visitors is on you. Most vendor widgets ship disclosure-ready — most businesses just never turn it on. That's a five-minute fix we find for you.
Is this legal advice?
No. It's a documentation and checklist service: we check your disclosures against the published rules and write the fixes. If the memo surfaces something that needs a lawyer, it says so.
What if December 2 passes and I haven't done anything?
Then the grace period for synthetic-content marking is gone and the obligation is fully live. Fix it now while it's a $149 checklist instead of a panic.
December 2 is the deadline. August 2 already passed.
Email us your URL and we start the audit. Handled by email, no cart, no calls.
Check my disclosures