IAB introduces an AI-era attribution blueprint
The Interactive Advertising Bureau (IAB) has released a new framework intended to address one of digital advertising’s most consequential unanswered questions: when artificial intelligence helps a consumer discover, choose, or buy a product, who should receive credit for the sale?

The AI-Era Attribution Blueprint, published on November 12, 2026, does not claim to solve attribution completely. Instead, it establishes a common vocabulary and evidence-based process for classifying AI’s role in a conversion and deciding whether a brand, agency, publisher, AI platform, retailer, or measurement provider can make a credible claim for credit. The IAB describes it as a supplement to existing attribution systems, not a replacement for them.
Why the industry needs it
Traditional digital attribution has long relied on observable signals: a click on an advertisement, a referral from a website, a search query, or a direct visit to a retailer. Those signals allow marketers to connect an exposure or interaction with a later purchase.
AI changes that model. Consumers may ask an AI assistant for product recommendations, receive a suggested item, compare options, and then complete a purchase without ever clicking a conventional ad or visiting a publisher’s site through a trackable referral. In such cases, the sale may be recorded as search traffic, direct traffic, or another measurable channel, even though AI materially shaped the decision.
The IAB says this is already a practical measurement problem. In its 2026 Outlook Study: September Update, 45% of ad buyers identified comparing AI-driven and traditional customer journeys as their top measurement challenge. The same research found that 44% rely on branded search and direct traffic as proxies for AI influence they cannot directly track.
That creates a risk of misallocation. If an AI recommendation drives interest or product selection, but the eventual purchase is attributed to direct traffic, the AI platform may appear less valuable than it is. Conversely, if AI platforms receive credit without sufficient evidence, advertisers may overstate their impact. The blueprint is designed to reduce both types of error by making the basis for any credit claim explicit.
What the blueprint asks
At the point of sale, the framework asks three questions:
Who or what created the shopper’s intent?
Who or what made the purchase decision?
Who or what completed the transaction?
Those questions determine whether a conversion is classified as Human-Owned, AI-Assisted, or AI-Executed.
Human-Owned conversions
A conversion is Human-Owned when the person develops the intent, chooses the product, and completes the purchase. AI may still be involved—for example, by answering a factual question—but it does not influence what the shopper wants or selects.
In this pathway, the human remains the primary decision-maker. The framework therefore treats the outcome differently from a sale in which AI introduced a product or steered the consumer toward a particular choice.
AI-Assisted conversions
A conversion is AI-Assisted when AI influences the consumer’s interest or product choice, but the person completes the purchase. This can include an AI system introducing a brand into consideration or recommending a specific product that the shopper accepts.
This category is likely to be the most commercially important because it covers a broad middle ground: the consumer still makes the final purchase, but AI has affected the consideration set, product selection, or both. The blueprint recognizes that influence does not automatically equal credit; a party must meet the applicable evidence standard before its claim can be evaluated.
AI-Executed conversions
A conversion is AI-Executed when an AI agent completes a purchase using permission granted in advance. The consumer may have specified exactly what to buy, or may have allowed the agent to select a product within defined parameters.
This pathway raises a distinct attribution issue because the human may not actively select the product at the moment of purchase. The framework therefore separates the act of authorizing an agent from the act of choosing a product, and requires evidence appropriate to the role each party played.
Claims are not credit
One of the blueprint’s most important distinctions is that a claim is not credit. Under IAB’s language, a claim is “the right to be evaluated for credit.”
This matters because AI-driven commerce creates multiple potential claimants. An AI platform may argue that its recommendation generated the purchase. A publisher may argue that its content informed the consumer. A retailer may argue that its storefront completed the transaction. A brand may argue that its advertising created prior awareness. The blueprint does not automatically award credit to any of them; instead, it defines which parties are eligible to make a claim and what evidence each claim requires.
The framework distinguishes between a decision claim and a contributory claim. A decision claim relates to the party or system that shaped the purchase decision, while a contributory claim relates to a party that contributed to the outcome without necessarily making the final decision. The applicable pathway—Human-Owned, AI-Assisted, or AI-Executed—determines whether either type of claim is available and what proof is needed.
The measurement gap
The blueprint is candid about a central limitation: much of the evidence needed to substantiate AI-related claims is either not captured or is held within AI platforms.
This is not a minor technical issue. If an AI system recommends a product, the advertiser may not be able to see the prompt, the sources used, the ranking logic, whether the recommendation was paid or organic, or how the consumer interacted with the recommendation. Without those details, it is difficult to verify whether AI merely answered a question, introduced a brand, changed the consideration set, or effectively made the purchasing decision.
Reporting on the release noted that limited access to measurement data—including how AI generates recommendations and affects consumer decisions—prevented the IAB from publishing more robust criteria for assigning credit to AI search platforms. The hardest distinction, according to the reporting, is separating AI’s role in shaping intent from its role in shaping the final decision.
The IAB does not present the blueprint as a finished measurement standard that resolves every dispute. Rather, it identifies the transparency and measurement gaps that currently prevent claims from being substantiated and asks the industry to help close them.
What each part of the ecosystem must do
The IAB assigns different responsibilities to different participants.
Brands are asked to estimate the share of their conversions that fall into each AI pathway. This would give advertisers a clearer view of how much of their business is Human-Owned, AI-Assisted, or AI-Executed, rather than treating all conversions as though they followed the same path.
AI platforms are asked to disclose whether recommendations are paid or organic and to identify the sources that informed them. They are also asked to label AI-originated traffic consistently. These disclosures would make it easier for advertisers and measurement providers to assess whether an AI recommendation was commercially influenced and whether it can be connected to a conversion.
Agencies are asked to apply a single claim-and-evidence standard across platforms. They are also expected to separate platform-reported claims from independently verified ones. That distinction is important because a platform’s own report of influence is not necessarily independent proof that it caused a sale.
Publishers are asked to identify the contributory claims their content creates and the evidence needed to support them. This acknowledges that editorial content, reviews, comparisons, and other publisher materials may still influence purchase decisions even when an AI system is the immediate interface.
A practical example
Consider a shopper who asks an AI assistant for a recommendation for a specific type of product. The assistant provides several options, explains the differences, and highlights one product. The shopper then visits the brand’s website directly and completes the purchase.
Under conventional attribution, that sale might be recorded as direct traffic. Under the IAB framework, the relevant question is whether the AI recommendation influenced the shopper’s interest or product choice. If it did, the conversion may be AI-Assisted rather than Human-Owned. The AI platform may then be eligible to make a claim, but the claim would need supporting evidence; it would not automatically receive credit merely because a purchase followed a recommendation.
If instead the shopper authorized an AI agent to buy a defined product without further human selection, the transaction could be classified as AI-Executed. The evidence requirements and eligible claimants would differ accordingly.
What it builds on
The blueprint was developed by the IAB AI-Era Attribution Blueprint Working Group, a cross-industry group involving brands, agencies, publishers, AI platforms, and measurement providers. It builds on IAB’s earlier Measuring Visibility in the AI Era playbook, extending the focus from how brands appear within AI platforms to how that presence connects to sales.
IAB CEO David Cohen framed the initiative as an acknowledgment of what the industry cannot yet substantiate, while also inviting ecosystem participants to apply the framework and improve the observability, verifiability, and measurability of AI’s role in commerce.
Caroline Giegerich, IAB’s Vice President of AI, said that as AI takes on more of the purchase journey—from sparking interest to completing sales—the clicks and referrals on which attribution depends are thinning or disappearing. She argued that AI-influenced sales are sometimes reported as search or direct traffic, potentially sending future marketing investment to the wrong channel.
Why it matters
The blueprint is significant because it moves the discussion beyond whether AI influences commerce and toward how that influence should be measured, verified, and allocated. It gives the industry a shared classification system at a time when AI-mediated discovery is becoming more common but the underlying data remains unevenly distributed.
Its practical effect will depend on adoption. Brands, agencies, publishers, retailers, AI platforms, and measurement providers will need to agree on what evidence is sufficient, how AI-originated traffic should be labeled, and how conflicts between competing claims should be resolved. The IAB has said it will bring these groups together to share learnings as more data becomes available and to continue narrowing the measurement gaps.
For now, the blueprint’s central contribution is clarity: it separates human intent, AI influence, AI decision-making, and transaction execution; it defines who may make a claim; and it insists that a claim must be supported by evidence before it can be treated as credit.
Sources
IAB — “IAB Releases AI-Era Attribution Blueprint,” November 12, 2026.https://www.iab.com/news/iab-releases-ai-era-attribution-blueprint/
IAB — “The AI-Era Attribution Blueprint,” November 12, 2026.https://www.iab.com/guidelines/the-ai-era-attribution-blueprint/
IAB — IAB AI-Era Attribution Blueprint, November 2026 (PDF).https://www.iab.com/wp-content/uploads/2026/11/IAB_AI-Era_Attribution_Blueprint_Nov_2026.pdf
IAB Canada — “Who gets credit when AI influences a purchase?,” November 19, 2026.https://iabcanada.com/who-gets-credit-when-ai-influences-a-purchase/
Digiday — “The IAB finds AI measurement standards elusive amid data gaps,” November 12, 2026.https://digiday.com/media/media-briefing-the-iab-finds-ai-measurement-standards-elusive-for-publishers-amid-platform-data-gaps/
eMarketer — “IAB hits roadblocks for AI attribution framework, revealing extent of measurement gaps,” November 18, 2026.https://www.emarketer.com/content/iab-hits-roadblocks-ai-attribution-framework-revealing-extent-of-measurement-gaps
ExchangeWire — “Digest: IAB Unveils Framework for AI Purchase Attribution,” November 16, 2026.https://www.exchangewire.com/blog/2026/11/16/digest-iab-unveils-framework-for-ai-purchase-attribution-radio-leads-on-ad-attention-relative-to-carbon-emissions/
MediaPost — “IAB Releases AI-Era Attribution Blueprint,” November 12, 2026.https://www.mediapost.com/publications/article/419316/iab-releases-ai-era-attribution-blueprint.html
SUMMARY
1. IAB released the AI-Era Attribution Blueprint on November 12, 2026, to standardize how the industry classifies AI’s role in consumer purchases and evaluates claims for credit.
2. It categorizes conversions as Human-Owned, AI-Assisted, or AI-Executed, while distinguishing between a claim and verified credit.
3. The framework highlights major measurement gaps, particularly limited access to AI-platform data, and calls for greater transparency from brands, agencies, publishers, and AI platforms.



Comments