The Trust Premium

(Was this newsletter forwarded to you? Sign up here.)

Gartner's name for the shift this spring was "a new scarcity. Not attention, but trust."

It's a strange sentence, and worth a second read. For the entire history of marketing, the scarce inputs were the obvious ones. Reach was scarce, so we paid for distribution. Production was scarce, so good creative was a moat. Attention was scarce, so we fought for it. AI just made two of those three effectively free. Anyone can produce competent creative at volume now, and the reach to push it is a rounding error. The thing that didn't get cheaper, the thing that actually got rarer as the feed filled with machine-made content, is whether anyone believes you.

The consumer data backs it up, hard.

• 78% of consumers say they'd rather see ads made by people, even if AI could produce better ones (Canva, May 2026).

• 63% say ads in AI search results would make them trust those results less (Ipsos, January 2026).

• 75% say they'd trust an AI shopping agent less if its recommendations were swayed by brand dollars (Quad and The Harris Poll, February 2026).

• Only 7% say AI-made marketing increases their trust in a brand. 31% say it decreases it. 91% expect to be told when something is AI-made (Klaviyo and Datalily, March 2026).

Two wrong lessons

Most of the industry is drawing exactly the wrong lesson from this, in two opposite directions.

The first wrong lesson is the maximalist one: production is free now, so flood every channel. That's the move that manufactures the distrust the surveys are measuring. The second wrong lesson is the reactionary one: ban AI, slap "human-made" on everything, treat it like a moral stance. iHeartMedia's Guaranteed Human initiative gets quoted as the model here, and the soundbite is good ("we're on the side of humans"), but "no AI anywhere" is a position that gets more expensive and less credible every quarter AI gets better.

The right read is in the second half of iHeart's own line, the part nobody quotes: "automate the pipes, but never the personality." That's not a slogan. That's an operating principle, and it's the most useful thing anyone has said about AI and trust this year.

Trust is not a property of your whole operation. It lives at specific points: the recommendation, the voice, the judgment call, the moment a customer decides whether to believe what you just told them. Everywhere else, in the rendering, the resizing, the data cleanup, the first draft, the thousand variations, AI is pure upside and the customer neither knows nor cares. The mistake isn't using AI. The mistake is using it at the point where trust is actually transacted.

The Pipes and the Personality

So draw the line on purpose. Take every place AI touches your marketing and sort it into one of two buckets.

The pipes are everything the customer experiences the output of but never the authorship of. Production, versioning, media buying, analytics, the operational plumbing. Automate the pipes aggressively. This is where the efficiency case for AI is real and the trust cost is zero, because trust was never being transacted there.

The personality is everything that carries your judgment to the customer: the recommendation you stand behind, the brand voice, the point of view, the human who is visibly accountable for what's being said. Spend trust capital here and protect it. The moment a customer suspects the personality is automated, the discount applies, and the studies below show how steep it is.

Here's the part that makes this structural rather than a style preference. The discount doesn't attach to AI quality. It attaches to AI disclosure. A peer-reviewed study in June 2026 found that simply labeling identical content as AI-made cut how much effort people believed went into it. The Nuremberg Institute found the same ads rated less natural and less useful purely because they were labeled AI. And 91% of consumers now expect that label. You cannot out-quality your way past this, because the penalty is triggered by the disclosure, not the craft, and the disclosure is becoming mandatory. That's why "just make better AI creative" doesn't solve it, and why protecting the personality is a durable position rather than a nostalgic one.

Won't the premium evaporate?

The strongest pushback is that this is temporary. Human-made is a quality signal today only because AI is still occasionally bad and obviously labeled. Once AI creative is reliably excellent and labeling fades into background noise, the premium evaporates and the people who built a brand on "made by humans" look like they bet on the buggy whip.

It's the right objection, and the evidence cuts against it. If the premium were about quality, better AI would erase it. But the studies isolate the label from the output: same content, worse perception, purely because of the disclosure. That's not a quality gap closing over time. That's a provenance preference, and provenance preferences are sticky. We still pay more for handmade, for single-origin, for "made in" a place, long after the machine-made version became objectively fine. The trust premium isn't a bet that AI stays bad. It's a bet that people keep caring who, or what, is behind the thing they're being asked to believe. That bet has a long winning record.

Where to draw the line

Draw the pipes-and-personality line explicitly, in writing, with your team. Most organizations are making this call by accident, tool by tool, with no one deciding where the line sits. Decide it on purpose. Here is where we automate without hesitation, and here is where a human stays visibly accountable. The line is a strategy, not a setting.

Get ahead of disclosure instead of getting caught by it. 91% already expect to be told. Regulation is heading the same way, and a funded campaign is spending up to $8 million to push it there faster. Brands that disclose on their own terms, framed around where they deliberately keep humans in the loop, control the story. Brands that get disclosed onto, by a regulator or a journalist or a sharp-eyed customer, don't.

Audit where you're spending trust to save pennies. Somewhere in your operation, AI has crept from the pipes into the personality because it was cheaper, and you're quietly paying for it in a currency that doesn't show up on the efficiency dashboard. Find the recommendation, the response, the voice that got automated to save a marginal cost, and ask whether the savings were worth the trust. Usually they weren't, and usually no one ran the math.

Reach got free. Production got free. Trust didn't, and it won't, and that's the whole opportunity. The brands that win the next few years won't be the ones that used the most AI or the least. They'll be the ones that knew exactly where to stop.