Labelling AI in Advertising: Disclosure, Platform Rules and Trust

Short answer: rules on disclosing AI in advertising are still developing, but the direction is clear: realistic synthetic content, especially people who appear real, or events that did not happen, increasingly needs to be labelled. Brands should follow each platform’s current policy, watch regulation in their markets, and adopt a simple, honest disclosure practice of their own.

What is changing

• Regulation: the EU AI Act includes transparency obligations for certain AI-generated or manipulated content (so-called deepfakes), phased in over time. Other jurisdictions are developing their own rules.
• Platforms: major video and social platforms ask creators and advertisers to disclose realistic altered or synthetic content, and some apply labels automatically.

Details differ by market and change frequently, check current guidance or legal advice for each campaign.

What usually needs disclosure

• Realistic people who do not exist, presented as real customers or experts.
• Real people depicted doing or saying things they did not do.
• Realistic scenes presented as documentary or real events.

Clearly stylised, fantastical or animated content is generally treated differently.

A practical policy

Keep a production note for every campaign listing where AI was used; label realistic synthetic people and events; never use a real person’s likeness without written consent; and make testimonials real.

Why it helps

Audiences are increasingly alert to AI imagery. Being open about it, and using it where it adds creativity rather than deception, protects trust. See AI commercial production.

Frequently asked questions