The Interactive Advertising Bureau (IAB) published its AI-Era Attribution Blueprint on November 12, 2026. The framework classifies purchases by who sparked the shopper’s interest, who made the product choice, and who completed the order. Those roles help determine who can make an attribution claim and what evidence would support it, according to IAB’s announcement.
A merchant may see a purchase recorded under search or direct traffic even when an assistant helped the shopper decide. The blueprint follows that path to the sale, building on IAB’s August guidelines for measuring visibility in AI answers.
IAB classifies a purchase by looking at who or what sparked the intent to buy, who chose the product, and who placed the order. The blueprint defines three pathways and splits AI-assisted purchases into two types based on how the assistant influenced the shopper.

The source of the product choice matters even when an assistant handles checkout. An order record may show that an assistant placed the order. Evaluating its role in the decision may require seeing whether the shopper named the product or asked the assistant to select it.
An AI system that determines the final product choice may make a decision claim. A party whose content or data informs the recommendation may make a contributory claim. The party evaluating the claim weighs the evidence before awarding credit. The criteria are set out in IAB’s published blueprint.
Referral data can show that a shopper followed a link from an AI answer. A retailer’s transaction record can confirm the sale. Connecting those events becomes harder when shoppers return through another channel. In a June study of ChatGPT recommendations in finance, travel, and beauty, 55.9% of visits to recommended brands in the following week came through search, according to Similarweb’s analysis.
IAB calls on AI platforms to say which sources informed a recommendation, indicate whether it was paid or organic, and label traffic that originates in their tools. It also asks agencies to distinguish platform-reported claims from those they can verify independently.
Suppose a shopper asks an AI assistant to recommend a carry-on bag that meets an airline’s size limits. The assistant cites an affiliate publisher’s comparison guide and recommends a model. The shopper later searches for the model by name and buys it from the retailer. In that AI-assisted purchase, the citation shows the guide appeared in the answer. The merchant may record the visit as a search and see the purchase in its order data. The publisher would need evidence linking those events to substantiate its contribution.
Publishers can compare the AI referrals they can identify with outbound affiliate clicks and credited sales over the same period. The AI Assistant channel can help identify those referrals, as explained in Affiverse’s guide to tracking AI traffic in GA4. To reconcile a commission on a specific order, the program manager needs a record that ties the purchase to the publisher under the program’s attribution rules.