Sixty-two percent. According to a survey Optimizely published last month, 62% of consumers now trust marketing content less since brands started using AI. In the same report, 67% find AI-generated human imagery in ads “creepy.” The industry reads this number as an aesthetics problem — fix the visual, improve the prompt. It isn’t. It’s an economics problem, and its economics has already been described.
Brandolini’s Bill
The principle named after Italian programmer Alberto Brandolini is simple: the cost of producing bullshit is an order of magnitude lower than the cost of refuting it. A decade ago this principle lived mostly in debates about journalism and disinformation. Today it’s every marketing department’s daily reality. Producing a campaign line, an image, a statistical claim takes seconds; verifying that it’s accurate, consistent, on-brand still takes human time — and as production speeds up, verification doesn’t speed up with it. Its queue just gets longer.
The strange part is how invisible this stays on the budget. An organisation boasts about generating a hundred variations, but nobody asks who approved that hundred, in how much time, against what standard. Verification isn’t a production step; it’s a separate discipline that begins where production ends — and precisely because it’s separate, it has no line of its own, no owner, no allotted time. Work with no line item unsurprisingly doesn’t get done.
Where the Abundance Moved
I’ve written before that when production gets cheap, scarcity doesn’t disappear — it moves to the next box over. This report gives that move an address. What’s grown scarce isn’t content anymore; it’s trust in content. The consumer isn’t complaining “I’m not hearing from this brand”; they’re saying “I don’t believe what I hear from this brand.” The difference looks small, but it isn’t: the first is a reach problem, solvable with budget; the second is a reputation problem, and budget doesn’t touch it.
The sly part of this shift is how late it shows up on the dashboard. Click-through, impressions, engagement — all keep behaving normally for a while, because trust erosion stays quiet at first: people don’t leave right away, they just keep reading a little less convinced. The bill arrives the day the brand makes a claim that’s actually true — and nobody believes it, because the muscle for believing has already been weakened for months.
Who Pays for Verification
Most organisations still don’t bill this to anyone — verification is a job that’s supposed to “somehow get handled,” with no official owner. My proposal is a new role on the production line: before anything goes live, someone other than the person who made it, working on a different clock than production speed, questions the claim and the image from scratch. That costs something, I know. But as Brandolini put it, not asking always costs more — the bill just arrives late, and late bills always arrive larger.
This role doesn’t need to be expensive, and it doesn’t need to be slow. The only requirement is independence: the person who made the piece can’t approve their own work, because speed is already on their side — verification’s whole job is finding the one person who can stand against it.
I’ll leave two questions, precisely because they’re hard to answer:
- How much of what you published last month did someone other than the maker read start to finish before it went live?
- Is the cost of rebuilding trust lower or higher than the cost of never losing it?
Few marketers know the answer to the second. Maybe that’s where the real scarcity is.