Why is more AI making your marketing worse

Speak your PR description, bug reproduction, or Cursor prompt. Wispr Flow auto-tags file names, preserves variable names, and formats everything for immediate paste into GitHub, Jira, or your editor.

No re-typing. No context gaps. No mangled syntax. Works natively inside Cursor, Warp, and every IDE at the system level.

4x faster than typing. 89% of messages sent with zero edits. Used by engineering teams at OpenAI, Vercel, and Clay.

Try Wispr Flow free

There is a quietly uncomfortable fact sitting underneath every marketing team's growth dashboard right now.

The brands using the most AI are not the ones growing the fastest.

This is going to sound counterintuitive given everything you have read on LinkedIn for the last 18 months. Every newsletter, every podcast, every founder thread has been some variation of the same story. Use AI to write more emails. Use AI to spin up more landing pages. Use AI to generate more social posts. Use AI to scale your content team from three people to thirty. The narrative has been so consistent and so loud that even careful operators have started to feel like they are falling behind if they are not pushing AI deeper into their marketing operations every quarter.

But the data coming out of consumer research over the last six months tells a different story. A story that almost no marketing team is ready to confront, because confronting it would mean rolling back a lot of recent decisions.

The story is this. Consumers have crossed a threshold. They can now spot AI-generated content with reasonable accuracy. When they spot it, they reduce their engagement, their trust in the brand, and their willingness to buy. The brands going harder on AI right now are not gaining a velocity advantage. They are accumulating a slow trust debt that does not show up in this quarter's CAC numbers but will show up in next year's retention numbers.

This edition is about that gap. Why is it happening? What the actual numbers look like. And what the brands quietly winning right now are doing that almost everyone else is missing.

Let me put the actual research on the table because I think the magnitudes matter.

A 2025 study from the Nuremberg Institute for Market Decisions surveyed 600 marketing professionals and found that 100% of them were using AI in their work. Not 80%. Not 95%. 100%. AI is no longer a competitive advantage in marketing. It is table stakes.

On the consumer side, the numbers move in the opposite direction. Roughly half of consumers can now correctly identify AI-generated content when they see it. When they identify it, 52% report reduced engagement. 62% are less likely to engage or trust content on social media if they know it was generated using AI. 26% find AI-generated website copy impersonal. 31% of consumers say AI in ads makes them less likely to choose a brand.

A 2026 survey reported that 54 %of Americans are experiencing what researchers are now calling "AI fatigue," a measurable drop in tolerance for the volume and quality of AI-driven communication.

And the gap between marketers and consumers on this point is the largest in the entire study. 77% of advertisers view AI in marketing positively. Only 38 %of consumers do.

What this means in plain language is that the marketing industry has collectively decided that AI is the future of how brands communicate, while the people those brands are trying to communicate to have collectively decided they do not want that future. Both groups are operating on different planets right now, and only one of them is the customer.

The trust gap is not theoretical. It is showing up in measurable ways inside the funnel of every brand that has gone heavy on AI in the last 18 months. You can see it if you know where to look.

The first place it shows up is in cold email response rates. A year ago, AI-personalised cold email was producing meaningfully higher reply rates than generic outreach. The novelty of the personalisation cut through. Today, those same sequences are producing reply rates that are flat or worse than templated emails from three years ago. Recipients have learned to recognise the cadence of AI-generated personalisation. The "I noticed your recent post about X" opener that worked in 2024 now triggers an immediate delete because the recipient has had hundreds of identical emails. The personalisation has become its own pattern, and the pattern is now a negative signal rather than positive.

The second place is in landing page conversion rates. Brands that have spun up dozens of AI-generated landing pages for different audience segments are quietly seeing their conversion rates compress over time. The pages look optimised. They have all the best conversion practices. They were generated faster than any team could have written them. But they convert worse than the original handwritten page they replaced, because they read like AI to the visitor, and the visitor's trust is now affected before they even get to the offer.

The third place is in social engagement. Look at the engagement curves on any LinkedIn account that ramped up AI-assisted posting in 2024. Total impressions have stayed roughly constant or even grown. But engagement rate, the percentage of impressions that actually do anything, has dropped meaningfully. The audience is still seeing the content. They are no longer responding to it. Because the content reads as automated, and audiences no longer reward automated content with engagement, the way the algorithms are still rewarding it with reach.

The fourth place is in email open rates and click-through rates over time. Brands that automated their lifecycle emails with AI templates often see strong initial performance, then a slow degradation over six to twelve months. The recipients have not unsubscribed. They are still on the list. They have just learned that emails from this brand are not worth opening, and they have stopped opening them.

In every case, the surface metrics, total volume of content, total impressions, and list size often look fine or improved. The engagement metrics, the ones that actually predict revenue, are quietly worse than they were before AI was scaled.

You might fairly ask why this trust collapse is happening now. AI-generated content has been around for two years. Why is the rejection showing up only in the last six months?

The answer comes down to a tipping point in pattern recognition. AI-generated content has visual, structural, and rhythmic tells. The em dash usage. The "it is not just X, it is Y" construction. The bulleted lists have a parallel structure. The em-rule of three in titles. The inevitable closing paragraph that ties everything together with a slight emotional uplift. The way every sentence is roughly the same length. The smooth professionalism does not have any genuine roughness anywhere in the writing.

For the first 18 months of widespread AI content, consumers were exposed to it but had not yet pattern-matched it. They knew something felt off without being able to articulate why. They might disengage from a piece of AI content, but they did not know that was the reason.

In the last six to nine months, the pattern recognition has become explicit. Articles in the New York Times, the Atlantic, and dozens of consumer publications have explicitly named the markers of AI writing. Social media has surfaced the patterns. People now consciously notice and explicitly name AI content when they encounter it. The em dash has become a meme. "ChatGPT wrote this" is a common reply to LinkedIn posts.

Once the pattern becomes conscious, the behaviour shifts. People do not just disengage. They actively distrust. The content does not just fail to land. It actively reduces the credibility of whoever published it.

This is the threshold marketing teams have crossed without realising it. The same content that was effective in 2024 is now actively harmful in late 2025 and 2026, because the audience has changed, even though the content has not.

Here is what makes this particular moment so dangerous for marketing teams.

The smart response to this trust collapse is asymmetric. The brands that will win in the next two years are the ones that use AI heavily for internal operations, for analysis, for research, for first-draft acceleration, but who reduce visible AI in customer-facing communication. They are using AI to think faster while writing more carefully, not the other way around.

But the dominant industry narrative is pushing marketing teams in the opposite direction. Use AI to scale content. Use AI to generate more variants. Use AI to automate more of the customer-facing surface area. The advice that sounds intuitive is actively wrong.

This creates a window where the brands paying close attention to consumer behaviour can pull ahead, specifically because their competitors are scaling the wrong way. Every brand racing to put more AI-generated content in front of customers is making it easier for the brands that quietly choose human authorship to feel different. The contrast is the differentiation.

If your competitor is publishing 50 AI-generated LinkedIn posts a week and you are publishing five carefully written ones with real, specific stories from your experience, your five posts now stand out more than they would have in 2022. The signal-to-noise ratio has flipped. AI-generated content has become the noise. Genuine human writing has become the signal that breaks through.

This is the strategic opening that almost no one is exploiting because it requires going against the loudest voices in marketing right now.

The brands I am watching grow well right now are not anti-AI. None of them is taking a public stand against AI or making it part of their positioning. That would be performative, and the audience can read performative.

What they are doing is more disciplined.

They are using AI heavily inside the building. For research synthesis. For analysing customer interview transcripts. For surfacing patterns in support tickets. For first drafts of internal documentation. For generating multiple variations of an idea so the marketer can pick the strongest direction and then write the actual final piece by hand. AI is making its internal velocity faster and its decisions sharper.

They are using AI cautiously outside the building. The customer-facing surface area, the emails, the landing page copy, the social posts, and the founder content are mostly written by humans, often by founders themselves, with AI used at most for editing assistance or specific structural suggestions. The voice is recognisably human. The specificity is real. The stories are first-hand.

They are leaning into the things AI cannot do. Original opinions. Direct, specific stories from their own experience. Strong takes that depend on having lived something. The kind of writing where the writer is genuinely on the line, where the argument might be wrong, where the personality is visible in the prose. AI cannot do these things, not because the model is incapable, but because the model has no personal stake in being right or being wrong. It cannot have a strong view because it has no skin in the game. The brands that lean into this are creating content that AI literally cannot replicate, which is exactly the moat in a flooded market.

They are slowing down content velocity and increasing content density. Instead of publishing five posts a week, they are publishing two posts a week with real research and real specificity. Instead of sending three emails a week, they are sending one email a week that is genuinely worth reading. The volume is lower. The engagement per piece is dramatically higher. The relationship with the audience is stronger.

This is a deliberately countercyclical strategy. While the rest of the industry is producing more, faster, and cheaper, these brands are producing less, slower, and more carefully. The audience is responding to the contrast.

If you want to evaluate where your brand sits on this curve, here is the simple test.

Print out the last 10 customer-facing pieces of content your brand published. Emails, posts, landing page copy, ad creative. Read them out loud as if you were a stranger encountering them for the first time. Ask three questions about each one.

First, does this piece read like it was written by a human who knows something specific? If you cannot identify a moment of specific knowledge, real experience, or genuine personality in the piece, your audience cannot either, and the piece is doing less work than you think it is.

Second, would I genuinely send this to a smart friend if it were not promotional? The smart friend test is brutal but useful. If you would not send the content to a friend because it would feel embarrassing or generic, that same feeling is what your audience experiences when they read it. They will not consciously articulate it. They will just engage less.

Third, could a competent competitor produce something basically identical with the same AI tools? If yes, the piece has zero strategic value, regardless of how well-written or optimised it appears. The point of marketing content is to be something only your brand could have produced. The moment that test fails, you are competing on volume, and volume is the most expensive game in marketing.

The brands that are quietly winning right now are the ones whose content passes all three tests. The brands that are quietly losing are the ones whose content passes none of them, and whose teams have not yet noticed because the surface metrics still look fine.

The shift is uncomfortable to make because it requires walking back decisions that felt right when they were made. But the shift is what separates the brands that will still be growing in 2027 from the ones that quietly stalled out without understanding why.

Concretely, three actions.

Audit your customer-facing content for AI markers. The em dashes. The "it is not just X, it is Y" sentences. The smooth professionalism without a single specific detail. The closing paragraphs wrap everything up too neatly. Then rewrite the worst offenders by hand.

Reduce content velocity by half. Whatever you are publishing, cut it in half. Use the time saved to make the remaining half twice as specific, twice as personal, and twice as honest. Volume is cheap right now and getting cheaper. Specificity and honesty are getting more valuable every quarter.

Have the founder write again. The single highest-leverage thing most founder-led brands can do right now is have the founder return to writing customer-facing communication directly. Not approving drafts that someone else wrote. Actually writing. The signal of a real founder voice, in this market, is one of the strongest differentiation signals available, and it gets stronger every month as more brands automate.

The dominant marketing narrative right now is going to look obvious in retrospect. The brands that won the AI era will not be the ones that used it most. They will be the ones who know where to use it and where not to.

That choice is sitting in front of every marketing team right now. Most are going to get it wrong because the path of least resistance is to follow the loudest voices on LinkedIn. The few that get it right, quietly, deliberately, and against the consensus, are going to look like the obvious winners three years from now.

What is in front of you is the opportunity to be one of those few.

See you at the next edition, Arindam