AI disclosure rules for marketers and creators
Marketers and creators must disclose AI use when audiences could be misled without it. In the EU, Article 50 of the AI Act has required deepfakes to be labelled since 2 August 2026. In the UK there is no blanket rule, but the ASA expects AI content not to mislead. In the US, FTC rules ban fake AI reviews and deceptive endorsements. In India, realistic AI-generated audio, images and video must be labelled since February 2026. Platforms such as Meta, TikTok and YouTube also require labels for realistic AI content.
What is the basic rule?
Across markets, the common principle is simple: do not mislead people about what is real. If a reasonable viewer could think an AI-generated person, voice, event or product result is genuine, and that would change how they understand the content, disclose it. Using AI to brainstorm, edit copy or retouch a background usually does not need a label. Creating a realistic spokesperson, a cloned voice or a scene that never happened usually does.
What does the EU require?
Article 50 of the EU AI Act has applied since 2 August 2026. People who deploy AI to create deepfakes, meaning realistic image, audio or video content that resembles real people, places or events, must disclose that it is artificially generated. For clearly artistic, satirical or fictional work, the disclosure can be made in a way that does not spoil the work. AI-generated text published to inform the public on matters of public interest must also be labelled unless it has had human editorial review. Chatbots must tell people they are AI. See the European Commission’s Article 50 FAQ and our EU AI Act guide.
What does the UK require?
There is no blanket legal requirement in the UK to label AI in ads. The Advertising Standards Authority applies its existing codes regardless of how content is made. Its guidance suggests asking whether the audience would be misled if AI use were not disclosed, and whether disclosure would clarify the message. A label cannot fix an ad that is misleading at its core, such as an AI “after” image showing results a product cannot achieve. The Digital Markets, Competition and Consumers Act also bans fake reviews, including AI-written ones.
What does the US require?
The Federal Trade Commission’s rule on fake reviews and testimonials, in force since 2024, bans reviews and testimonials that misrepresent who wrote them, including AI-generated reviews presented as from real customers. The FTC’s Endorsement Guides require endorsements to reflect real experiences and material connections to be disclosed. Some states have rules on AI in political ads and on digital replicas of real people’s voices and likenesses.
What does India require?
Amendments to India’s IT Rules, effective 20 February 2026, require realistic AI-generated audio, images and video to be clearly labelled and carry provenance metadata where feasible, and large platforms must ask users to declare AI content. AI-generated text alone is excluded. The Advertising Standards Council of India has also issued draft guidance on labelling AI-generated ads.
What do platforms require?
| Platform | What to do |
|---|---|
| Meta (Facebook, Instagram, Threads) | Use the AI label for realistic AI-generated or altered content; ads may show “AI info” |
| TikTok | Turn on the AI-generated content label for realistic AI content |
| YouTube | Use the altered or synthetic content disclosure for realistic content |
| Content may show provenance information from C2PA metadata |
Platform rules change often; check each platform’s help centre before a campaign. Our guides to AI avatar video tools and AI video ads cover common formats.
What about creators and influencers?
- Disclose paid partnerships as usual, with platform tools or #ad
- Label AI-generated versions of yourself or virtual influencers where viewers could be misled
- Do not present AI-generated product results as real experiences
- Get written permission before creating a likeness or voice of a real person
- Keep records of how content was made
Our UGC content guide covers disclosure for creator content more broadly.
What should be on a marketing team’s checklist?
- Identify AI elements: List any AI-generated people, voices, scenes, text or reviews.
- Apply the mislead test: Would a viewer think it is real, and would that matter?
- Label clearly: Use visible text or platform labels, not buried small print.
- Keep provenance metadata: Do not strip C2PA or tool metadata from files.
- Check permissions and rights: Likeness, voice and copyright.
- Log decisions: Record what was AI-generated and why you did or did not label it.
Rights and ownership questions are covered in AI video copyright, and our AI regulation comparison covers the wider legal picture.
What does this look like in practice?
A cosmetics brand runs campaigns in the UAE, UK, EU and India. Its agency creates an AI-generated model for product videos. The team labels the videos “Created with AI” on screen and turns on platform AI labels, keeps C2PA metadata and avoids showing AI-generated skin results. Customer testimonials come only from real customers with records. One label standard covers every market, and the agency logs every AI element.
How UIUX Studio helps
We help teams build AI content workflows with disclosure built in. Talk to our Agentic AI & Automation team. For compliant campaigns and creator content, Social Mafia by UIUX Studio runs influencer and paid programmes. This article is general information, not legal advice.
Frequently asked questions
Do I have to label all AI-generated content?
No; label it when people could be misled about what is real.
Is AI labelling required in the UK?
There is no blanket rule, but ads must not mislead.
What does the EU AI Act require?
Labels on deepfakes and some public-interest AI text since August 2026.
Can I use AI-generated reviews?
No; fake reviews are banned in the US and UK.
Does India require AI labels?
Yes, for realistic AI audio, images and video since February 2026.