The Interactive Advertising Bureau plans to publish a framework on Nov. 12 for measuring and crediting advertising exposure when AI agents, not humans, do the browsing and buying. The framework does not exist yet. It is still being drafted, and the people writing it cannot agree on the basics.
Caroline Giegerich, the IAB’s vp of AI, described the goal as building “a shared framework for measuring and crediting AI’s role in conversions, especially when traditional signals are minimized.” The problem she is trying to solve is concrete: when an AI agent reads a product page, compares it against competitors, and completes a purchase for a user, standard tracking such as UTM parameters and referral data often does not survive that path. Without those signals, publishers, platforms, and measurement vendors have no shared evidence for who influenced the sale.
Giegerich is drafting the document from conversations with a working group that includes tech companies, publishers, agencies, measurement vendors, and brands. She declined to name the participants. Asked what the group has struggled most to agree on, her answer was “everything.” Publishers in the group are pushing to be counted in attribution at all, arguing their content shaped the AI-generated response a user acted on.
That contested process, not a finished standard, is the actual news here. Search and marketing teams often treat working-group announcements as settled direction. This one is a starting point with a hard deadline attached, and the framework’s substance could shift considerably before November.
The IAB is expected to split AI’s role into two categories: cases where AI surfaces something to a user in an awareness or intent capacity, and cases where AI helps that user make the purchase decision. That split matters for search teams because it determines whether a citation in an AI answer counts the same as an AI agent completing a transaction, two very different signals for a publisher trying to prove value.
Bombora’s global data partnerships head, Jaime Schultheis, framed the effort as a chance to rebalance a relationship she says has favored platforms. “Many big tech partners have grown big audiences off of publishers, and there has been very little reciprocation,” Schultheis said, calling the IAB’s work “a tremendous opportunity for it to be truly a reciprocal relationship.”
The open question is whether AI platforms will supply the evidence any framework would need. Michael Bishop, who co-founded the AI-native ad platform OpenAds, said a workable standard still has to define what is measured, who measures it, how, and at which layer. He warned that AI platforms could remain black boxes even with an industry standard in place. Bishop compared the risk to how measurement vendors operated inside Facebook: unable to run their own tracking tags, dependent on integrations Facebook itself coded, and reduced to manually testing that Facebook’s own numbers checked out. “The black box was basically maintained,” he said, and “it does not look very good for anyone.”
Giegerich acknowledged that some of the needed evidence simply does not exist today. The IAB’s stated role is to push the parties who could produce it toward that conversation, not to compel disclosure. For search and marketing teams, the near-term task is not adopting a new measurement standard. It is tracking whether the November framework produces enforceable disclosure commitments from AI platforms, or a definitional scheme with no data behind it.
Teams that report on AI-driven traffic and conversions should flag November 12 now and watch specifically for what evidence obligations, if any, land on AI platforms themselves.
Reported by Sara Guaglione for Digiday, published Aug. 24, 2026.
