Is AI-generated creative allowed on Meta ads?

Meta has no policy against AI-generated creative. Nothing in the Advertising Standards prohibits an ad image, video or piece of copy because software produced it, and Meta ships its own generative AI features inside Ads Manager, including background generation and text variations. Millions of ads running on Facebook and Instagram today contain AI-generated or AI-edited elements, and that is entirely within the rules. What Meta cares about is the finished ad, not the tool. Its Advertising Standards state plainly that "it is an advertiser's responsibility to understand and comply with our policies," along with local laws and regulations. The ad review process looks at the specific components of an ad, including images, video, text and targeting information, as well as the landing page it points to. An AI-generated image that implies personal attributes, shows prohibited before-and-after body transformations, or makes misleading claims gets rejected for exactly the same reason a photographed image would. The reverse is also true: a clean, honest AI image sails through review like any other creative.

There are two separate mechanics here, and advertisers often mix them up. The first is a mandatory self-disclosure, and it applies only to ads about social issues, elections or politics. For those ads, Meta requires you to disclose when the creative contains a photorealistic image or video, or realistic-sounding audio, that was digitally created or altered to depict a real person saying or doing something they did not, a realistic-looking person that does not exist, a realistic-looking event that did not happen, or altered footage of a real event. Meta's ad standards are explicit about enforcement: if it determines an advertiser has not disclosed as required, it will reject the ad, and repeated failure to disclose may result in penalties against the advertiser. If you sell skincare, food delivery or tuition classes, this disclosure regime does not apply to you. The second mechanic is labeling, and it can touch any ad. Since 2024, Meta has applied an "AI Info" label when it detects industry-standard AI indicators in media or when the person uploading discloses AI use. Meta's Transparency Center adds that from June 1, 2026 it uses automated detection technology to identify ad media created or edited with third-party generative AI tools, and applies the AI Info label when it finds them, with no action required from the advertiser. In practice this means a photorealistic AI product scene may carry a small informational label whether you flag it or not. The label is context for viewers, not a policy strike.

When AI-made ads get rejected or accounts get restricted, the cause is almost always a standards violation the tool made easy to commit, not the AI itself. Generative models will happily produce a slimmer body for a weight-loss ad, an exaggerated results claim, a fake celebrity likeness, or a product render that does not match what you ship. Each of those breaks an existing rule: personal health and appearance policies, misleading claims, or misrepresentation. AI also invents details confidently, so a generated pack shot can show ingredients, certifications or price tags that do not exist, which becomes your compliance problem the moment you click publish. The practical standard is simple: treat AI output as a junior designer's first draft, not a finished asset. Check that the product shown is your actual product, that any people depicted do not resemble real public figures, that claims in generated copy are ones you can substantiate, and that the image does not drift into a sensitive category. Tools that ground generation in your own brand assets reduce this risk, which is why AdPlay.ai generates ad images and copy from a brand kit rather than from a blank prompt, keeping logos, products and claims anchored to what your business really offers.

Before launching AI-generated creative on Facebook or Instagram, run four checks. First, category check: if your ad touches social issues, elections or politics, complete Meta's authorization flow and tick the digitally created or altered disclosure whenever photorealistic AI media is involved. Second, standards check: review the creative against Meta's Advertising Standards exactly as you would a photo shoot, looking at claims, depicted people, health and appearance implications, and landing page consistency. Third, honesty check: confirm the AI image faithfully represents the real product and offer, because misrepresentation rules apply regardless of how the image was made. Fourth, label readiness: assume the ad may carry an AI Info label, and make sure the creative still builds trust with that label attached, which usually means avoiding fake photorealistic testimonials or staged customer scenes and leaning on clean product-led visuals instead. Ads that pass those four checks are as safe to run as any other creative on Meta.