IAB Releases AI Transparency and Disclosure Standards V2
The IAB released Version 2 of its AI Transparency and Disclosure Framework on August 18, 2026. The update tells advertisers, agencies, publishers and platforms to disclose AI when it materially affects authenticity, identity or representation. It stops short of asking brands to label every use of AI in marketing.
What the IAB AI Disclosure Framework Requires
The framework sorts AI use into two groups. The first group warrants a label. That covers realistic synthetic content that could change what consumers believe they are seeing. Examples include images and videos generated from prompts, some synthetic voices and avatars, and digital twins of people who have died. It also covers digital twins of living people placed in fabricated situations outside a standard brand endorsement. Chatbots and assistants belong here too. The IAB says they should be identified as AI when a consumer could reasonably mistake them for a human representative.
The second group does not trigger automatic disclosure. That includes routine post-production, internal workflows, text and copy, standard audio enhancement, background music, generic synthetic voices, and avatars that are obviously cartoons or stylized. Some authorized synthetic voices and digital twins used in conventional endorsements also sit outside the automatic-disclosure line. The nature of the AI use matters more than the fact that AI was involved.
The IAB drew on consumer research it ran with Sonata Insights for the first version of the framework. That research found mixed views. Some consumers saw AI in creative work as a positive. Others called it inauthentic. More than half wanted brands to disclose when an ad was fully generated by AI or contained AI-generated imagery or video.
For US advertising, the framework offers two disclosure formats: a standardized sparkle icon or clear text. Both are IAB recommendations and do not override the law. The EU takes a different route. Article 50 of the EU AI Act requires disclosure for covered AI-generated content and deepfakes but does not mandate an icon. A voluntary Code of Practice covers design and placement, and a common EU icon remains unsettled.
Who Reported the IAB AI Disclosure Update
MarTech’s Constantine von Hoffman reported the update on August 18, 2026. The framework itself is published by the IAB as AI Transparency and Disclosure Standards V2. The IAB released the first version in January 2026.
How AI Disclosure Rules Fit B2B Marketing in 2026
AI disclosure now sits alongside a wider push to measure and govern AI in marketing. The IAB published a separate framework this year for tracking brand visibility inside AI answers, covered here when the IAB released its Measuring Visibility in the AI Era framework. Regulators are moving on a second front. Britain’s competition regulator ordered Google to let publishers opt out of AI search, a decision covered in the UK CMA’s world-first AI search ruling. Ad platforms are building AI surfaces at the same time. OpenAI has been testing multi-advertiser ads inside ChatGPT, which places brand messages inside an AI answer rather than beside a search result.
The regulatory timeline is tight. New York’s synthetic performer law took effect in June. California’s SB 942 and Article 50 of the EU AI Act both took effect on August 2, 2026. South Korea introduced labeling requirements earlier this year. A campaign running across those markets can face four sets of rules at once. The IAB is offering one decision process to sit above them.
Talking Shift: In B2B, the AI You Must Disclose Has a Face or a Voice
Start Some Shift reads this framework as easy news for most B2B teams and a warning for a few. Most B2B AI work is text: emails, ad copy, landing pages, sales decks. The IAB says that kind of work does not need a label. The disclosure question lands somewhere narrower. It lands on the synthetic spokesperson in your product video, the AI voice in your webinar recap, the chatbot sitting on your pricing page. Those are the assets a buying committee judges you on. Start Some Shift’s take: “In B2B, the AI you have to disclose is almost always the AI you put a face or a voice on.” A buyer who spots an undisclosed AI voice in your demo will start wondering what else you left out. Run the audit asset by asset.
What B2B Marketers Should Do About AI Disclosure Rules
- Audit every live campaign asset for AI-generated faces, voices and avatars, since those are what the IAB framework flags first.
- Leave routine AI copywriting and post-production unlabeled, which the framework does not treat as an automatic disclosure trigger.
- Map each campaign to the markets it runs in, since California, New York, South Korea and the EU now apply separate rules.
- Choose your US disclosure format now, either the IAB sparkle icon or clear text, so creative teams stop deciding case by case.
- Log what the AI actually did to each asset, because the framework turns on the type of use.
- Treat a visible disclosure as a reason to click, since your ad sits beside competitors making similar claims with no label at all.
- Brief legal before the next campaign build, because IAB guidance does not override the law in any market.
What to Watch Next on AI Advertising Disclosure
Watch whether the EU settles on a common disclosure icon, since a shared symbol would make cross-border campaigns far simpler to build. Watch also whether more US states follow California and New York, which would turn today’s patchwork into something closer to a national standard.
