Short answer: add an AI disclaimer when (1) people talk to an AI (chatbot, voice agent, AI email replies), (2) you publish AI-generated or AI-altered content that could pass as human-made or real, or (3) AI output informs advice or decisions about people. Say what the AI does, its limits, how to reach a person, and how data is handled. Draft yours with the free AI disclaimer generator.
This article is general information, not legal advice. Check the rules that apply in your country and sector.
When an AI disclaimer is required or expected
| Situation | Why disclose |
|---|---|
| Chatbots and voice agents talking to the public | The EU AI Act (Article 50) requires that people are told they are interacting with AI unless it is obvious. Several US state laws have similar bot-disclosure rules. |
| AI-generated images, video, audio or text presented as real | The EU AI Act requires marking of synthetic content and disclosure of deepfakes; platforms such as YouTube and Meta require labels for realistic AI content. |
| AI-assisted health, financial or legal information | Consumer protection and professional rules: readers must not mistake it for personal advice. |
| Automated decisions about people (hiring, credit, insurance) | GDPR Article 22 and similar laws give rights around solely automated decisions; many jurisdictions require notice. |
| Marketing and reviews | Regulators such as the US FTC treat undisclosed AI that misleads consumers, including fake reviews, as deceptive. |
For the EU timeline, including what the 2026 Digital Omnibus changed and what it did not, read our EU AI Act update.
What a good AI disclaimer includes
- That AI is involved, in plain words, at the start of the interaction or next to the content.
- What it does (answers questions, books appointments, drafted this article).
- Its limits: may make mistakes, not professional advice, check important details.
- How to reach a person.
- Data use: whether conversations are stored or used to improve the service, with a link to the privacy policy.
Examples you can adapt
Website chatbot
"Hi, I'm the ZetsApp virtual assistant, an AI. I can answer questions about our services and book a call. I can make mistakes, so please confirm anything important with our team. Type 'agent' at any time to reach a person. Chats are stored to improve our service; see our privacy policy."
AI voice receptionist
"Thanks for calling Bright Dental. I'm the practice's virtual assistant and this call may be recorded. I can book or change appointments, or put you through to the team."
AI-assisted article
"This article was drafted with AI assistance and reviewed, fact-checked and edited by [Name], [Role]. Last updated [date]."
AI-generated image
"Image generated with AI. It is an illustration, not a photograph of a real event or product."
Health, finance or legal information
"This information was prepared with the help of AI tools and is general in nature. It is not medical, financial or legal advice. Speak to a qualified professional about your situation."
Automated screening
"We use an automated system to help review applications. A member of our team reviews every decision, and you can ask for a human review or an explanation by contacting [email]."
Where to place it
- Chat and voice: in the first message or greeting, not hidden in the terms.
- Content: next to the content (caption, byline, top of article), plus metadata or watermarking where your tools support it.
- Products: in the interface where the AI acts, and in your terms and privacy policy.
Mistakes to avoid
- Burying the disclosure in the footer or terms only.
- Pretending the assistant is a named human employee.
- Using a disclaimer as a substitute for accuracy. "AI can make mistakes" does not excuse publishing errors you could have checked.
- Forgetting to update the privacy policy to cover AI vendors and data flows.
Frequently asked questions
Do I have to say my chatbot is AI?
In the EU, yes, under the AI Act's transparency rules, unless it is obvious from the context. Elsewhere it is increasingly required and is best practice everywhere; it also builds trust.
Do I need to disclose that I used AI to write a blog post?
Not in most places for ordinary text that a human reviewed and is accountable for. Disclosure becomes important when content could mislead (realistic synthetic media, news, reviews) or covers health, finance or legal topics.
Does a disclaimer protect me from liability?
Only partly. It sets expectations, but you remain responsible for what your AI says and does, especially in regulated sectors.
Should the disclaimer be in every language I support?
Yes. Show it in the language of the conversation or page.
Create yours
Use the AI disclaimer generator for chatbot, content and advice disclosures. Building an AI product? Our GDPR-compliant AI development checklist covers the data side.
