In January, IAB Canada covered the first version of IAB’s AI transparency framework. IAB has now released an updated framework with more specific guidance on when and how to disclose AI use in ads.
The framework still focuses on whether AI use could mislead someone about what is real. The update gives examples across images, video, audio and text. It recommends disclosure for uses such as realistic AI-generated imagery, synthetic people who could be mistaken for real people and AI voices that portray someone making statements they never made. Routine edits, such as colour correction or noise removal, generally do not call for a label.
The updated framework also sets out steps for advertisers to assess AI use before launch, record their decisions and check that disclosures appear wherever an ad runs. Advertisers remain responsible for the accuracy of their claims, whether or not AI was used.
The framework is voluntary guidance. Read the full document for examples by content type and advice on applying disclosures.