There is a certain circularity to modern advertising that would have appealed to Joseph Heller, had he lived long enough to see programmatic buying.
Here is how it works, according to Andrew Frank, a vice president and distinguished analyst at Gartner: When misinformation about a brand begins circulating on artificial intelligence answer engines — the assertion, say, that a company is engaged in fraud or that its products cause harm — people naturally begin searching for more information. They type queries like "is [brand] a scam" into Google or ChatGPT, as one does when confronted with an alarming claim at the breakfast table.
The brand's automated bidding systems, which have been trained to interpret any surge in brand-related searches as a signal of consumer interest, detect this activity and respond in the only way they know how: by bidding more aggressively to appear alongside those searches. Often, Mr. Frank noted, no human being is involved in this decision.
The result is that brands may find themselves spending considerable sums to place their advertising directly adjacent to false narratives about themselves — narratives that the advertising, by its presence, may inadvertently lend credibility to. (One imagines the marketing budget as a dog enthusiastically chasing its own tail, except the tail is on fire.)
This phenomenon represents a new wrinkle in the long-running tension between the efficiency that automation promises and the judgment that it cannot provide. For decades, advertisers have sought to remove human hands from the machinery of media buying, on the reasonable theory that machines can process more information faster and cheaper. What the machines cannot do, it turns out, is recognize when the information they are processing is poisoned.
The solution, presumably, is more human oversight — which is to say, the very thing the automation was meant to eliminate.
Original story published in adweek.com: "Brands May Be Bankrolling the AI Misinformation Spreading About Them"