The IAB is making moves to tackle one of the thorniest problems in digital advertising today: how to measure, attribute, and give credit to ads being delivered not to people, but to artificial intelligence agents.
It sounds like something out of a sci-fi episode, but it is already happening. Ads are starting to be served to bots instead of humans, and nobody agrees on how much that is worth or who should get credit for the exposure. The Interactive Advertising Bureau wants to sort out this mess before it becomes impossible to fix.
Think about it: if an AI agent reads a product page, compares it with competitors, and completes a purchase on behalf of the user, who gets credit for that conversion? The publisher who created the content that shaped the AI response? The platform that ran the agent? Or simply nobody, because the evidence trail does not even exist yet?
That is exactly the gap the IAB wants to address with a measurement and attribution framework set to be published on November 12, 2026. The traditional signals the industry has always relied on to track conversions, like UTM parameters and referral data, simply do not survive the user journey inside AI platforms. And without reliable signals, there is no way to know if an ad actually contributed to a sale. The problem runs deeper than it looks. 👇
What Does It Mean to Advertise to an Agent?
According to Caroline Giegerich, VP of AI at the IAB, the goal is to create a shared framework for measuring and crediting the role of artificial intelligence in conversions, especially when traditional signals are stripped down to almost nothing. The organization is exploring new measurement mechanisms and signals to quantify AI influence on purchase decisions.
But opinions across the industry are far from aligned. As Giegerich herself put it, one side may see this as an interesting test, while the other side sees it as pure smoke and mirrors. She is drafting the framework now, drawing on conversations with a working group made up of tech companies, publishers, agencies, measurement vendors, and brands. Giegerich declined to reveal which companies are part of the group.
Receive the best innovation content in your email.
All the news, tips, trends, and resources you're looking for, delivered to your inbox.
By subscribing to the newsletter, you agree to receive communications from Método Viral. We are committed to always protecting and respecting your privacy.
Underneath it all, there is a desire to figure out what comes next in a market that is being turned upside down right now. In her words, the IAB is trying to keep up as much as possible with the flood of activity around the concept of agents that bypass traditional human traffic. It is a race against time, and nobody wants to be left behind.
The Problem Nobody Expected to Solve This Soon
For years, the digital advertising ecosystem was built on a simple premise: there is a human on the other side of the screen. That human clicks, browses, searches, adds to cart, and checks out. Every step of that journey leaves trackable digital breadcrumbs — cookies, session identifiers, campaign parameters, and HTTP referral data. The entire ad measurement infrastructure was built on top of that premise. The problem is that this premise is starting to crumble, and the industry has not yet realized the full extent of the damage it could cause.
Artificial intelligence agents do not browse the way humans do. They do not have persistent cookies, they do not land on destination pages the conventional way, they do not click banners, and they do not generate trackable sessions in the way analytics tools recognize. When an AI agent searches, compares prices, reads reviews, and finalizes a transaction, it does so in an environment that is completely different from what the advertising industry was designed to observe. The signals that do exist in this process are fragmented, opaque, and most of the time invisible to the attribution platforms that brands and agencies rely on today.
This behavioral shift is not something coming down the road — it is already happening. As these agents become more sophisticated, the trend is for more and more stages of the purchase journey to be delegated to them. The advertising industry needs a fast response, and that is where the IAB steps in with a structured proposal.
What the IAB Framework Aims to Do in Practice
The biggest challenge the framework will face is connecting AI visibility with the measurement of attribution for ads served to intelligent agents, while also understanding how AI is influencing marketing as a whole. The idea, according to Giegerich, is to separate AI impact into two distinct categories: when AI serves something to a user, acting as a layer of awareness or intent, and when AI actually helps that user make a decision.
Worth noting that this is not the IAB’s first move in this direction. The organization has already released frameworks focused on visibility measurement in the AI era and on disclosing the use of artificial intelligence in content production. The new document takes things a step further, specifically targeting the attribution question.
The core proposal is to establish technical and conceptual standards that make it possible to identify when an AI agent was exposed to advertising content, how that content influenced the decision the agent made, and how credit should be distributed among the participants in the chain. To achieve this, the framework needs to go beyond traditional attribution models like last click or multi-touch, because those models assume a human journey with clearly trackable touchpoints. In the world of AI agents, the logic is different: the agent may have been influenced by content it processed internally, without leaving any visible external signal.
The Big Standoff: Who Provides the Evidence?
The biggest obstacle so far has been exactly how to assign credit and what evidence is needed to do so, especially considering that AI platforms and tech companies are not exactly known for sharing that kind of information. As Giegerich put it, part of that evidence simply does not exist. And if it does not exist, but there is a party that could potentially provide it, then that is a good open discussion to have.
This dynamic could help build a better foundation in the relationship between publishers and tech platforms. That is the view of Jaime Schultheis, head of global data partnerships at Bombora, where she leads a network of B2B publishers, brand sites, and data providers. According to her, many major tech partners have built massive audiences on the back of publisher content with very little reciprocity. For Schultheis, this is a tremendous opportunity to make the relationship truly reciprocal.
But it is not going to be easy. When asked what the hardest thing has been to get all parties in the working group to agree on, Giegerich answered with a single word: everything. Publishers, in particular, argue that their content fed the response the AI system is serving to the user, and they do not want to be left out of the attribution conversation in any way.
AI Platforms Are Still Black Boxes
An AI advertising measurement framework would need to answer some fundamental and, honestly, pretty tough questions: what is being measured, who does the measuring, how is it done, and at which layer does it happen. That is according to Michael Bishop, co-founder of the native AI advertising platform OpenAds.
Bishop raises an important warning: even if the IAB manages to put together an industry standard that answers some of these questions, AI platforms may continue to be black boxes. If tech companies like OpenAI control the signals that show how AI influenced marketing outcomes, external measurement firms could end up relying on integrations or data provided by the platforms themselves — something very similar to what happened in the past.
He draws a direct parallel with Facebook, which built and controlled integrations with vendors to deliver measurement data without ever giving full visibility into its own systems. As Bishop explained, measurement vendors on Facebook did not run their own JavaScript tags. Facebook coded the integration to grade its own homework, and vendors essentially tested manually to rubber-stamp that integration. The black box was maintained, and vendors effectively acted as a trusted third-party audit layer. In his assessment, that kind of arrangement does not sit well with anyone.
Why This Discussion Matters Right Now
For advertisers, publishers, and digital marketing professionals, the practical impact of this framework could be massive. Today, when a campaign underperforms, there is at least an imperfect but traceable data chain that allows you to diagnose the problem. In the world of artificial intelligence agents, that chain nearly vanishes. A product could be getting recommended or ignored by AI agents based on criteria the brand does not even know about, and there is no analytics report showing it happening.
With a framework established by the IAB, the expectation is that platforms will start creating integration points that allow some level of tracking within agent environments. This could include agent-specific session identifiers, metadata flagging when content was processed by AI on behalf of a human user, and logging protocols that record which content sources influenced a decision. None of this exists in a standardized way today, but these are exactly the elements a solid attribution framework would need to cover to be useful in practice. 🎯
For publishers, the issue is especially urgent. If an AI agent uses editorial content from a site to support a product recommendation that results in a sale, who gets any kind of compensation for that? Today, the honest answer is that nobody guarantees anything. The IAB framework could set the first precedents for how value generated by content within artificial intelligence environments should be recognized and distributed.
The framework expected in November 2026 represents the first serious, structured attempt by the industry to create clear rules for measuring and attributing the value of ads in an ecosystem where artificial intelligence agents are active participants in the purchase journey. It will not solve everything at once, but it puts the industry on the right track.
