Mockingbird

Learning Center

AI and ad creative

Advertisers now make ad images and copy with generative AI, in third-party tools and in Meta's own Ads Manager, where Advantage+ creative offers text generation, image generation, background generation and image expansion. The tools differ in how much of the real product and the brand survives into the finished ad.

This topic explains how AI ad generators work, the errors they make, and what Meta requires when an ad is made with AI, including the disclosure rule for political ads and the AI info labels Meta adds itself.

Key terms

Common questions

Does Meta have its own AI ad generator?
Yes. Advantage+ creative in Ads Manager includes generative AI features such as text generation, image generation, background generation for catalogue ads and image expansion to fit more placements. Features that use AI show an AI logo under their name, and some are on by default, so check the previews before you publish.
Is it safe to use AI-generated people in ads?
It is allowed, with conditions. Meta labels ads made with its tools that include an AI-generated photorealistic person, and political or issue ads must disclose realistic people who do not exist. Never generate a real or famous person's likeness without permission, because the UK Advertising Standards Authority warns that fake endorsements mislead and Meta runs facial recognition against scam ads that use public figures.
Can Meta tell if an ad was made with AI?
Sometimes. Meta reads industry-standard signals that AI tools write into files, and it named the C2PA and IPTC standards when it began labelling AI images in 2024. In June 2026 it said it would start using such signals to label ads made with third-party AI tools. Files without the signals, or with them stripped out, can go undetected.

Put it to work

Other topics