I asked an AI chatbot last month which small Greek island had the fewest tourists and the best hiking. It gave me three names, I picked one, booked a ferry, and spent five days wandering around Amorgos with barely another traveler in sight.
Here's the question that stuck with me afterward: if a hotel or tour company had paid for placement in that AI's answer, would anyone ever know it influenced my decision? Probably not. And that's exactly the mess the advertising industry is trying to untangle right now.
The Interactive Advertising Bureau just released something called the AI-Era Attribution Blueprint. It's a new framework meant to answer a question that didn't even exist a few years ago: when AI helps sell something, who gets credit for the sale?
A Problem Hiding in Plain Sight
For decades, advertisers tracked a simple trail. You click an ad, you land on a page, you buy something. Clean, traceable, attributable.
AI breaks that trail. If I ask a chatbot for restaurant recommendations in Lisbon and then just walk into one, there's no click. No referral link. Nothing for a marketer to point to and say "that's where the sale came from."
According to IAB's own research, this is already a real headache. In its 2026 Outlook Study, 45% of ad buyers said comparing AI-driven customer journeys to traditional ones is their single biggest measurement challenge. Even stranger, 44% admitted they're using branded search and direct website traffic as stand-ins for AI influence they can't actually track.
Think about what that means. A sale that an AI tool quietly nudged along might get logged as "direct traffic," and a company could end up pouring its next ad budget into the wrong channel entirely, chasing a signal that was never the real cause.
Caroline Giegerich, IAB's VP of AI, put it plainly: AI-influenced sales showing up as search or direct traffic can send marketing money in the wrong direction. That's not a small accounting error. That's a business making decisions based on a story that isn't true.
Sorting Out Who Did What
So what does the Blueprint actually do? It asks three questions about any purchase: who or what sparked the interest, who made the decision, and who completed it.
Based on the answers, every conversion gets sorted into one of three buckets: Human-Owned, AI-Assisted, or AI-Executed. That classification then determines who's even eligible to claim credit, and what proof they'd need to back it up.
I like one detail here. A claim isn't the same as credit. It's just the right to be evaluated for credit. That's a more honest starting point than most industries bother with. Nobody's declaring victory before the evidence is in.
The Evidence Problem
Here's the catch: a lot of that evidence doesn't exist yet, or it's locked inside the AI platforms themselves. So the Blueprint calls on everyone in the chain to pitch in.
Brands are asked to estimate how much of their business falls into each category. AI platforms are being asked to disclose whether recommendations are paid placements or organic results, and to label AI-driven traffic consistently instead of burying it. Agencies are expected to apply one consistent standard across platforms rather than trusting whatever each platform reports. Publishers are asked to figure out what role their own content plays in influencing a sale, and what proof supports that role.
IAB CEO David Cohen said something I appreciated: the industry has never had a complete picture of how influence turns into a sale, and AI just adds another layer to a problem that was already unsolved. Rather than pretending otherwise, the Blueprint is upfront about what still can't be proven.
Why This Matters Beyond Boardrooms
This isn't just a marketing-department problem. It shapes what gets recommended to you, what gets funded, and what disappears because nobody could prove it worked.
I didn't book that ferry to Amorgos because of an ad. I booked it because a tool I trusted gave me a good answer. Multiply that by millions of searches, chats, and recommendations happening every day, and you start to see why an entire industry needs a shared way to make sense of it.
AI isn't just changing how we travel, shop, or decide where to eat dinner. It's changing who gets to take credit for helping us decide. That's worth paying attention to, whether you're running a business or just asking a chatbot where to go next.
Source: Interactive Advertising Bureau
Related reading from us: Meta, TikTok and Google Just Told Indonesia: Slow Down on AI Copyright
