Should you tell people when content was made with AI? Increasingly, the law and the platforms decide part of the answer for you; your brand decides the rest.
Where disclosure is required or expected
- EU AI Act, Article 50: deployers of AI that creates "deep fakes" — realistic image, audio or video of people, places or events — must disclose that it was artificially generated or manipulated. For evidently artistic, creative or satirical work the duty is lighter. It applies from 2 August 2026 under the Act as adopted.
- New York: a law signed on December 11, 2025 (effective June 9, 2026) requires ads that include a "synthetic performer" — a digital human that isn't a recognisable real performer — to disclose it conspicuously, with civil penalties of $1,000 for a first violation and $5,000 after.
- Platforms: YouTube requires creators to disclose realistic altered or synthetic content (since March 2024) but not AI used for scripts or captions. TikTok automatically labels AI content carrying C2PA "Content Credentials" (since May 2024).
When small print isn't enough
In August 2025, Guess ran ads in Vogue featuring AI-generated models made by an AI agency. The ad carried a small credit line, and the campaign still drew criticism online about "fake" models and lost work for real ones. Disclosure meets a legal minimum; it doesn't settle the brand question.
Three stances a brand can take
- Label all substantially AI-generated content. Simple and transparent; may over-label routine work.
- Label realistic synthetic people, voices and scenes, plus anything required. A common middle path.
- Label only where required. Lowest effort; needs constant tracking of laws and platform rules by market.
How to label well
Plain words ("Made with AI"), close to the content, readable at the size people see it. Keep the content credentials that tools add; don't strip metadata. Apply the same rules to influencers and partners.