I, for one, am sick to death of hearing how brilliant AI is for creating marketing campaigns. Because for the most part, it isn’t.
New research from Meltwater and YouGov, based on nearly 10,000 consumers across seven markets, shows that a lot of people agree with me. While they aren’t rejecting AI completely, they’re watching how companies use it like a hawk, and trust in even established brands is now being eroded as a result.
I’ve written before about AI being a double-edged sword for marketers, and this report basically hands me a 10,000-person dataset that says “yep, you were right”. So naturally, I’m going to be smug about it.
Here’s what the research found, and why your business should consider using real people instead of generative AI.
People aren’t nearly as excited as you’ve been told
AI dominates every conference agenda, product roadmap, and LinkedIn thought-leadership post going, so it’s easy to assume the public shares that energy. Spoiler – they don’t.
Across the study, 39% of global consumers said they were excited about a future shaped by generative AI, while 51% felt uncertain or sceptical.
While younger audiences are noticeably more up for it, with 48% of 25 to 34-year-olds reporting excitement versus just 31% of the over-55s, the market splits are pretty bonkers. Germany (56%) and Singapore (55%) sit at the optimistic end, while the UK (23%) and the US (25%) are the most sceptical. So, if you’re producing AI-generated content for a British audience, you need to take note that not even a quarter of the people you’re talking to feel positive about the technology you are using to push marketing messages at them.
The fear isn’t AI itself, it’s what it lets people get away with
Here’s the bit I find most telling. When consumers say they’re worried about AI, it’s all based around trust: fake news, scams, misleading rubbish and the slow erosion of authenticity all rank high, and these concerns are held by big majorities in every single market studied. Misinformation in particular is a dominant worry, with 73% flagging it as a concern.
We’ve all watched AI slop flood the internet, and people are also becoming more aware of deepfakes and rage-bait that it can be used to create. So when your brand rocks up using AI, people don’t see the cost and resource savings that sold you on it in the first place. They see risk. And unless you tackle that head-on, it quietly chips away at your credibility, whether your particular use of AI deserves it or not.
There’s a cracking contradiction buried in here, too. Most people are pretty confident they can personally spot AI content, with 58% reckoning they can identify it. But an overwhelming majority worry that everyone else will be fooled.
So you’ve got a population that feels personally clever but socially uneasy, convinced the world around them is stuffed with deception. For brands, that means trust is fragile by default.
Even accurate, well-meaning content is landing in an environment where people already expect to be misled. You only need to look at the average comments left under a news story on Facebook to see how little trust there is in what people see online – even when presented with facts.
Where you use AI matters more than whether you use it
One of the clearest signals in the whole report is that acceptance depends massively on context. People don’t apply a blanket rule.
Acceptance is comparatively high in entertainment (53%) and advertising (47%), which are spaces where creativity, a bit of experimentation, and a degree of manipulation are expected anyway. Shift into higher-trust territory, and attitudes change in a heartbeat. Thankfully, only 21% find AI acceptable in news, and 18% accept it in politics (though I’d still argue this is too high). Influencer content sits fairly low, too, at 28%, which is pretty damning when you consider that the entire world trades on perceived authenticity. While AI influencers aren’t particularly new, this might make brands rethink whether or not to work with one.
So the question was never “should we use AI”, it’s “where does AI make sense for us, and where would it actively cost us trust”. A brand using AI to knock up a visual for a social media post is (generally) able to get away with it. But the same brand using AI to fake what looks like genuine customer testimony, making ads to promote products which don’t back up the claims made, or to churn out content dressed up as independent editorial, is asking for trouble.
AI ad scams
Here’s the thing the report can’t capture in a percentage, but every one of us is living through it: the absolute tidal wave of AI-generated scam ads currently clogging up YouTube, Facebook, TikTok, and Instagram. And these are actively training people to distrust advertising full stop, which is a problem for every legitimate brand running paid media.
Because we’ve just had a heatwave in the UK, I have seen about a million ads with a crappy AI voiceover and questionable product shots spliced into shitty b-roll on almost every bloody YouTube ad I’ve had to endure this week.
One that stood out (for all the wrong reasons) was a “portable air cooler” which pops up under a dozen different names, but as far as I can tell, it’s the same fan with a different crappy website. The ad is almost entirely AI-generated with a voiceover that spins this lovely yarn about how it was invented by a clever young engineer from Manchester. Heartwarming. Patriotic. Plucky underdog taking on ‘Big Air Conditioning’. And all complete bollocks.
I was interested to see where the ad went. Click the ad, and you don’t land on a shop; you land on a fake review site that has conveniently crowned this particular brand the #1 air cooler of 2026 with a score of 9.8, “1,938 customer reviews”, and glowing testimonials from suspiciously enthusiastic people in London, Manchester, and Edinburgh who use yank terminology. The “reviewer” is a chap called Eric Bates who, surprise surprise, has an AI-generated headshot and doesn’t appear to exist. The single listed con is “often out of stock due to high demand”, which is the sort of thing you only write when you’re trying to manufacture urgency rather than review a product. Buried at the bottom, in tiny text, is an admission that the site takes payment from the companies it “rates” and that compensation affects the rankings.
Then I went and looked at the actual reviews on Trustpilot. Real customers describe a glorified USB desk fan, checkout pages that silently duplicate your order so you pay for four units when you wanted one, warranty and shipping fees bolted on without consent, sneaky recurring payments, support emails that bounce, near-impossible refunds, and in at least one case, a unit that overheated and melted its own switch. The fake review site has its own Trustpilot page too, where people who twigged the con have left reviews to warn others.
This is AI-generated content, AI-generated imagery, and AI-fabricated social proof all working together to separate people from their money. And here’s why it matters even if your brand would never dream of doing any of this: every person who gets stung by one of these, or who simply learns to recognise the format, gets a little more cynical about the next ad they see. Yours included.
When 73% of people are already worried about AI-driven misinformation, this is the stuff fuelling it. The scammers are actively degrading the trust environment that legitimate advertisers have to operate in, which is exactly why authenticity matters (or at least disclosure of AI usage in marketing materials if you are going down that route).
AI starts at a trust disadvantage
When asked how learning that content was AI-generated would change their view of a brand, 32% said they’d trust it less, against only 15% who said they’d trust it more.
The efficiency gains and cost savings are real, but they’re an internal benefit, and something your customers don’t care about. Externally, AI use kicks off from a deficit you’ve got to claw your way back from. And it gets worse when certain lines get crossed: content that feels misleading, AI used with no disclosure, or AI that visibly replaces human creativity altogether. We saw exactly this with Coca-Cola and McDonald’s getting roasted for their AI Christmas ads.
The brands turning “no AI” into a selling point
Here’s the flip side, and it’s the bit I find genuinely interesting. If using AI starts you at a trust deficit, then loudly refusing to use it can be turned into an actual asset. A handful of brands have worked this out, and the early results suggest it pays off.
The obvious poster child is Dove. Back in 2024, it became the first major beauty brand to pledge it would never use AI to represent real women in its advertising, as an extension of its 20-year Real Beauty platform. The clever bit isn’t the pledge itself, it’s that consumers loved it. Ad-testing firm System1 ran the campaign through its emotional response measurement and rated it the best Dove ad they had ever tested, with audience happiness peaking at the exact moment the “keep beauty real” pledge lands on screen.
And Dove isn’t alone. Anti-AI marketing trends are beginning to take hold, with other major brands running “human-made” campaigns.
I feel like we’re at the point where the flood of AI slop has made genuinely human-made work into a way companies can stand out by demonstrating that they actually value their customers and have faith in the products or services they are selling. Pew found that 76% of people think it’s extremely or very important to be able to tell whether something was made by a human or by AI. NielsenIQ found that even AI ads judged to be high quality create weaker memory activation than human-made ones, so they literally don’t stick in the brain as well.
So the human angle isn’t just a feel-good ethical stance, it can be the more effective creative choice. That being said, plenty of brands will use AI and simply be more honest about where and how.
And I do genuinely understand that small businesses with limited budgets or creative skills now have a tool at their disposal to save both time and money to knock up images for social media posts and ads, which is not necessarily a bad thing (though I am also sick to death of local pubs using the same AI-generated slop to show off their summer menus).
Transparency is now the baseline, not a nice-to-have
A whopping 86% of consumers say AI-generated content should be disclosed, and that expectation holds firm across every market in the study. There is no region where people are relaxed about you hiding it.
For years, brands could focus almost entirely on outcomes, on whether the content was good, useful, or persuasive. Now the process matters too. People want to know how something was made, not just what it says, and the faintest whiff that you’re concealing something erodes trust fast, even when the concealment was completely innocent.
This is really just E-E-A-T wearing a different hat: trustworthiness is doing the heavy lifting, and disclosure has quietly become a credibility signal in its own right. Saying “we used AI here, and here’s why” is no longer the biggest risk. But hiding it is.
So what do you actually do about it?
The use of AI in marketing materials isn’t ending any time soon. Audiences know full well it’s being used more and more, but they are judging brands that use it.
The brands that will get away with it will be the ones who use AI deliberately, disclose it clearly, and keep a human firmly in the loop.
I’d advise clients to be picky about where AI adds genuine value instead of trying to use it for everything just because you can (or because shareholders want some more delicious cash, and sacking a creative team seems like the easiest way to achieve that).
Set a clear disclosure standard and apply it consistently, especially in the contexts where people are most sensitive. And actually monitor how your audience responds rather than assuming they’ll just accept it.