Table of Content
Meta is updating how AI-generated content is disclosed in ads on Facebook and Instagram, giving advertisers and users a clearer way to identify promoted posts that include synthetic or heavily AI-edited visuals.
The change places AI disclosure information inside the “About this ad” section, which users can access through the three-dot menu on promoted posts. This section is becoming Meta’s central destination for ad transparency, sitting alongside existing explanations such as why a user is seeing a particular ad.
The update comes as Meta adds more generative AI tools to its advertising system and prepares for tighter transparency rules in major markets. For advertisers, the message is clear: AI-made visuals are becoming easier to produce, but they are also becoming harder to hide.
How the New Label Works
Meta’s updated system applies an “AI info” disclosure when an ad includes images or videos created or significantly changed with generative AI. The label can be triggered in two main ways.
The first is when advertisers use Meta’s own AI ad tools. These include features that generate backgrounds, create images, or add animation to static creative. If the tool substantially changes an image or video, the ad may receive an AI disclosure.
The second path involves third-party AI tools. If an advertiser uses outside generative AI software to create or modify an ad, Meta can detect certain industry-standard provenance signals and apply the label when those signals are present.
The placement of the label depends on the level and type of AI use. Minor AI edits may not receive a visible label. More significant AI edits may be disclosed inside the “About this ad” menu. But if an ad includes an AI-generated photorealistic person, the disclosure can appear more prominently near the main paid-ad label.
That distinction matters. A synthetic background and a synthetic human are treated differently because audiences are more likely to be misled when a realistic person appears to exist, speak, endorse, or perform in an ad.
A Long Shift in AI Transparency
Meta’s approach to AI labels has changed several times over the past two years. Earlier labels were criticized for being too broad because some real photographs were marked as AI-assisted after only minor editing or cleanup. That created confusion for photographers and creators who used normal editing tools but had not generated fake scenes.
Meta later softened the wording and moved toward “AI info,” a broader phrase that can cover content made with AI or modified using AI. The latest ad update continues that direction. Instead of placing every disclosure directly in front of users, Meta is moving more information into a dedicated transparency menu while keeping more prominent labels for higher-risk cases.
This creates a balancing act. Meta wants users to know when ads include synthetic media, but it also does not want every small AI-assisted edit to make an ad look suspicious.
Why the Timing Matters
The update arrives as AI-generated ad creative is becoming a larger part of Meta’s business strategy. The company is building tools that let advertisers create variations faster, test different styles, change visual elements, and generate more campaign assets with less manual production.
Newer image-generation systems are also being added to consumer and business products, giving users and advertisers more ways to create AI visuals directly inside Meta’s ecosystem. That means Meta needs a disclosure pipeline that can scale with the amount of AI content its own tools produce.
The regulatory context is also important. New AI transparency obligations in Europe are set to take effect in August 2026, requiring clearer disclosure around certain AI-generated content and stronger machine-readable marking for synthetic outputs. Meta’s use of provenance signals and regional flexibility suggests the company is preparing for different disclosure standards across markets.
The Commercial Tension
AI labels create a real commercial tension for advertisers. Research and user feedback consistently show that people want to be warned when AI is used to create realistic content, especially when people appear to say or do things they never actually did.
At the same time, visible AI disclosures can affect trust. Some consumers may view AI-generated brand creative as less authentic, especially when it uses realistic people, influencer-style scenes, or synthetic endorsements.
That makes placement important. A disclosure hidden inside a menu and a disclosure placed beside the ad label are not the same user experience. The more realistic and human-like the AI output, the more visible the disclosure becomes.
What Is Not Covered
One important detail for creators, writers, and social media teams is that Meta’s ad framework is focused mainly on visual AI content. AI-written captions, bios, hashtags, usernames, and ordinary ad copy are not the main target of this labeling system.
That means a brand using AI to draft an Instagram caption or ad headline should not automatically expect a Meta AI label for that text alone. The system is aimed at generated or significantly edited images and videos, especially realistic media.
For businesses, the practical takeaway is simple. AI-generated visuals need more transparency planning, especially when they include realistic humans. AI-assisted text still needs normal legal, brand, and advertising review, but it is not treated the same way as synthetic visual media under Meta’s current ad-labeling approach.
Meta’s update shows where advertising is heading. Generative AI will be built deeper into creative production, but disclosure, provenance, and regional compliance will become part of the workflow. The brands that use AI well will not only create faster. They will also need to be clear about what was made, modified, and shown to users.