Meta signs the EU AI labeling code: an advertiser checklist

Meta confirmed on July 28, 2026 that it is signing the EU Code of Practice on Transparency of AI-generated Content. For advertisers the practical answer is: no new ad-specific deadline was announced on that page, and Meta already shows "AI info" labels on ads when its own generative tools make significant edits. Plan for labels to exist, and disclose accurately yourself.

What Meta's pages state

Facts below are from two Meta newsroom posts, plus the Commission's code page for context.

EU code context

The Commission's code page says the final code was published June 10, 2026, with about 190 signatories by the end of July 2026. Section 1 covers providers (marking and detection); Section 2 covers deployers (labeling). Advertisers using AI creative are closer to the deployer side, but read the code itself for your case.

Checklist

Nothing here is legal advice, and the Meta page gives no ad-specific steps, so this list is a working habit.

Sume's Formats can carry a repeatable ad workflow, but the docs do not say Formats add AI labels. Add the disclosure step to your own checklist.

• Log which creatives used AI tools and which were only lightly edited.
• Flag photorealistic AI humans; Meta's own labeling treats them differently.
• Look in About this ad on a live ad to see what a viewer sees.
• Do not assume Meta will detect and label everything; disclose where your rules require it.
• Re-read the Meta and Commission pages before each campaign. The July post gives no timeline.