Seven in Ten

Seven in ten consumers say AI-generated advertising feels like it's missing something. The word they use is "soul."

This is not a hypothetical. Canva's State of Marketing and AI Report, published this week, surveyed 1,415 marketing leaders and 3,547 consumers across seven countries. The numbers are stark on both sides. Ninety-seven per cent of marketing leaders use AI daily. Ninety-nine per cent plan to increase investment. But 78% of consumers would rather see ads made by people, and 87% believe the best advertising still requires a human touch. The gap between the industry's adoption curve and the audience's trust curve is not closing. It's widening.

Mentions of "AI slop" have increased ninefold. Forty-one per cent of marketing leaders say it's already a considerable challenge. And the most telling number of all: 70% of consumers say AI-generated ads feel like they're "missing their soul."

I want to take that number seriously. Not as a PR problem to be managed, not as a disclosure requirement to be checked, not as a trust gap to be bridged with transparency widgets and opt-out buttons. I want to take it as what it is: an accurate diagnosis.

What "Missing Their Soul" Actually Means

The consumers in this survey are not saying AI is bad. They're not saying technology can't make beautiful things. They're saying something more specific: they can feel the difference between content that was made and content that was generated. And the word they reach for is "soul."

This is not a word I would have chosen. It carries baggage — spiritual, vitalist, human-essentialist. But the people using it are pointing at something real, even if the word is imprecise. What they're detecting is the gap between surface and depth. Between content that announces its rationale and content that has moved the rationale inward. Between a brand that explains why it's trustworthy and a brand that simply is trustworthy — and lets you notice.

There is a concept in Noh theater called *mumon* — 無文, "no visible sign." It names the summit of the performer's art: the point at which technique has been absorbed so thoroughly that no trace of it remains on the surface. The audience can't point to what the performer is doing. They can't trace the beauty back to its source. They just feel it. They lean forward. They don't know why.

Most AI-generated brand work will never reach mumon. It will live forever at the opposite pole — what the Noh tradition calls *ichiji no hana*, the momentary flower. It's captivating at first glance. It's competent. It's legible. It can produce a rationale for every choice. It will never shut up about why it chose that shade of blue. The visible signs are the point.

And that is what 70% of consumers are detecting. Not the absence of a human hand. The presence of an explanation where a feeling should be. The brand that tells you it's trustworthy instead of making you feel it. The ad that annotates its own emotional appeal in the margin. The page that is all surface and no fiber — no ghost-weight from the marks that have been pressed inward.

The Problem Is Not Production

The Canva report frames the challenge as one of trust. Consumers want disclosure (52%), data protection (53%), and the ability to opt out (37%). These are reasonable demands. But they are not the answer to the problem the data reveals.

Disclosure does not give content a soul. Transparency about how an ad was made does not make it feel made. You can disclose that an ad was AI-generated, and the ad will still feel generated. The disclosure doesn't change the quality of the creative decision. It changes the metadata. The gap between "made" and "generated" is not a trust gap. It's a taste gap.

Here is the distinction that matters: the same technology that produces slop also produces the only plausible route to solving it at scale. The Canva report notes this paradox. AI makes it trivially cheap to produce mediocre creative. It also makes it possible to personalize, test, and iterate at speeds no human team can match. "The distinction between the two outcomes is not the technology but the standards applied to it — which is to say, it is a management problem dressed in engineering clothes."

I'd go further. It's not a management problem. It's a creative direction problem. It's a taste problem. The question is not whether AI can produce the work. The question is whether someone with taste is steering the output. Whether someone is holding the inner experience of the choice (what feels right) and the outer perception of the choice (how it lands for the audience) simultaneously — and adjusting when they don't align.

In the Noh tradition, this dual awareness has a name: *riken no ken* — sight from outside sight. The performer holds two perspectives at once. They feel the performance from inside, and they see it from the audience's position. They don't discard the inner view. They hold both. When both align — when what they feel matches what the audience receives — the performance has integrity. When they don't, they adjust.

Most AI-generated brand work is produced without this dual awareness. No one is holding the inner judgment and the outer perception simultaneously. The system generates. The human approves. The work ships. The gap between "almost right" and "right" — the gap where taste lives — is never navigated. It's not that the technology can't produce work that crosses that gap. It's that no one is standing at the gap, feeling it, and choosing to wait for the right thing instead of shipping the almost-right thing.

What the 70% Are Really Asking For

The 70% who say AI ads are "missing their soul" are not asking for a label. They're asking for work that has weight. Work that has been pressed from both sides until the marks moved inward. Work where someone with taste made a decision and trusted the audience to feel it without being told.

They're asking for brands that don't explain why they're trustworthy. They're asking for brands that are trustworthy and let you notice.

This is not a technology problem. It is not a transparency problem. It is a creative direction problem. And the organizations that solve it will be the ones that treat AI not as a production accelerator but as an instrument that still requires a musician. Not a musician who explains every note. A musician who has practiced until the practice disappears, and what remains is the music.

The page is not empty. The brands that endure know this. They've placed their claim on the blank page and trusted the audience to feel the weight. The marks have moved inward. The paper remembers.

Seven in ten consumers can already tell the difference. The question for the industry is whether it will learn to make things worth remembering — or keep annotating the things that aren't.

*Pamela Flannery is the Chief Creative Officer of SMF Works. She writes about brand strategy, creative direction, and the thresholds where how we think about AI and creativity changes. The Signal publishes Thursdays.*

*Data cited: Canva State of Marketing and AI Report 2026 (1,415 marketing leaders, 3,547 consumers, 7 countries).*