Open Instagram on any evening in October 2026 and count how many ads look like they came out of the same machine. Faces that are too smooth, hands with extra fingers, copy that promises to “transform your business” without saying how. People have noticed. And they are tired.
At the same time, it has never been harder to advertise without artificial intelligence. The two largest paid media platforms in the world are handing their algorithms decisions that, until recently, belonged to the advertiser: bids, audiences, copy and even visuals. For mid-sized companies, which need to grow but lack a marketing department the size of a multinational’s, this paradox has become the defining challenge of the year.
Fatigue, in numbers
In June, at the Cannes Lions Festival of Creativity, research by The Harris Poll with the 4As and Infillion showed the scale of the problem: 73% of consumers say they would trust an ad less if they suspected it was made with AI, and 63% say they would be less likely to buy from a brand that uses AI-generated ads. For 78%, AI makes ads feel less authentic. More than half (54%) say they are already annoyed by how much they hear about artificial intelligence.
European data points the same way. Klaviyo’s AI consumer trends report, based on 8,000 people in countries including Germany, France, Spain, Italy and the UK, found that visibly AI-generated marketing content makes 31% of consumers trust a brand less. Only 7% trust it more.
And it is not only small advertisers getting it wrong. In late 2025, Valentino was criticised for a video of handbags morphing into surreal shapes, which audiences described as “disturbing”. In the Netherlands, McDonald’s pulled an AI-made Christmas ad after it was called empty and sloppy. The term “AI slop” has firmly entered the vocabulary of anyone who talks about advertising online.
The platforms, however, are heading the other way
While audiences ask for less artificiality, the platforms ask for more trust in their algorithms. In mid-2025, The Wall Street Journal reported Meta’s plan to let brands create and target entire campaigns with AI by the end of 2026, simply by entering a product, a goal and a budget.
Google went further on paper. On 1 July 2026, it rewrote the Google Ads terms of service for the first time in about eight years, making automation the default and leaving advertisers responsible for reviewing, approving or removing whatever the system generates on its own. Since September, search campaigns using campaign-level broad match or automatically created assets have been moved to AI Max, with no opt-out.
The promised results have also been revised. During testing, Google said AI Max would deliver 14% more conversions. By the end of the beta, the published average had dropped to 7%. An independent Smarter Ecommerce analysis of more than 250 e-commerce campaigns, reported by the trade press, found a median effect of zero on return on ad spend and a 16% increase in cost per acquisition. On Meta’s side, a Business Insider investigation gathered cases of advertisers whose visuals were altered by the platform’s AI, such as a bicycle that gained a second set of handlebars. The company’s response was that verification rests with the advertiser.
None of this means automation is useless. Quite the opposite: properly set up, it finds audiences and combinations no human team would ever test manually. The problem lies elsewhere. Algorithms optimise for the goal they are given, and the platform’s goal is not always your business goal.
Why mid-sized companies are in the most sensitive spot
Large brands have in-house data teams, contracts with global agencies and budget to make mistakes. Small businesses often advertise little and sporadically. Mid-sized companies sit in between: they already invest meaningful sums in media, depend on qualified leads to grow and, precisely for that reason, are the most tempted by the promise of “autopilot”.
The reasoning seems logical. If the platform says AI optimises on its own, why not raise the budget and let the system work? Experience shows that more budget rarely fixes a poorly structured campaign. It just speeds up the waste. Without reliable tracking, AI learns from the wrong data. Without a clear goal, it optimises for what is cheapest to deliver, not what is most valuable to the company. Without creative with a real identity, the brand becomes just another generic ad in the feed, exactly the kind of content people have learned to ignore.
In Europe, there is an extra layer. GDPR privacy rules and Google’s Consent Mode v2 have changed how conversion data reaches the platforms. A poor setup means an algorithm that sees only part of the results and makes decisions blindly.
What separates companies that scale from those that just spend
Among the companies that manage to grow with paid media in 2026, a few points come up again and again:
- Structure before budget. Pixel, tags, consent and conversion events tested before adding another euro. This is the foundation automation learns from.
- One goal per campaign. Awareness, leads or sales. Mixing everything in one campaign confuses the algorithm and makes results harder to read.
- Audiences defined with intent. Automation broadens reach, but it needs a starting point: role, industry, buying stage.
- Metrics that matter. Cost per qualified lead and return on ad spend connected to the CRM, not just clicks, likes and reach.
- Human creativity as a differentiator. If audiences are saturated with machine-made ads, the piece that sounds human, specific and true becomes even more valuable. AI can speed up tests and variations, but the idea, the tone and the story are still the work of people.
- Scale only where the numbers prove it. Weekly tests of creatives, audiences and bids, with budget increases only where results hold up.
The role of a good marketing partner
This is where choosing a marketing partner stops being an operational detail and becomes a strategic decision. In a market where platforms automate execution, the value of an agency or consultancy lies less and less in “pushing buttons” and more and more in three functions no algorithm performs on its own.
The first is choosing the right channels. Not every company needs to be on every platform. Google Ads captures people who are already looking for a solution; Meta Ads builds desire, trust and brand recall; LinkedIn makes sense for B2B decisions. Defining each channel’s role in the funnel stops the budget from spreading without purpose.
The second is reading data with a business lens. A report full of charts is not the same as an analysis. The right partner cross-checks what the platform shows with what is happening in the company’s sales pipeline and can tell where the investment is leaking and where it is worth doubling down.
The third is acting as referee between the algorithm and the company. Knowing when to accept an automatic recommendation and when to turn it down, reviewing what AI generates before it reaches the public, protecting the brand’s identity. In some cases, the best results come precisely from not applying the most obvious optimisation the platform suggests.
There is also a transparency point that applies to any engagement: ad accounts, pixels and data should belong to the company, with full access and clear monthly reporting. A real partnership does not create dependency. It creates autonomy.
Less noise, more strategy
Artificial intelligence is not leaving digital advertising. It already sits at the heart of the platforms and will keep evolving. What is at stake for mid-sized companies is who steers the process: the algorithm alone, optimising for its own goals, or a clear strategy that uses technology as a tool and keeps people in charge of the decisions that really matter.
In a feed increasingly saturated with generic content, the competitive edge in 2026 is not advertising more. It is advertising better, with reliable data, well-chosen channels and messages that genuinely feel made for people.
Want to find out where your ad budget is leaking?
Agência Lira plans and manages paid advertising campaigns on Google Ads and Meta Ads for companies in Europe and Latin America, with diagnosis, strategy and weekly optimisation. Book your free ad audit.
Sources
- Marketing Brew: “Consumers are sick and tired of hearing about AI all the time” (The Harris Poll, 4As, Infillion), 06/2026
- EMARKETER: “Shoppers aren’t impressed by AI-generated marketing”, 05/2026
- Klaviyo: Consumer Trust in AI 2026
- McGraw Hill: “AI Slop Backfires in Ads”, 05/2026
- AIXH: “Performance Max, Advantage+, AI Max: who runs your budget?”, 2026
- Campaign Asia: “Meta aims to fully automate ad creation with AI by 2026”, 06/2025