The web is filling up with AI content. Google doesn’t seem to mind at all. People increasingly do.
There’s a genuinely strange split happening in digital marketing right now, and I don’t think it’s getting enough attention. On one hand, the algorithms that are deciding what gets ranked and cited appear largely indifferent to whether content was written by a person or a machine. On the other, the humans actually reading that content are growing sceptical.
Those two dynamics are in tension with each other and I think the businesses that understand the gap between them are going to have a genuine advantage over the ones that don’t.\
Does Google care where your content came from?
So what has changed? Ahrefs’ analysis of 900,000 new web pages found that 74% now contain some degree of AI-generated content, a figure that’s become the baseline rather than the exception in barely a couple of years (Ahrefs, 2025). Their follow-up research went further, looking specifically at the pages Google’s AI Overviews actually cite. It found that 87.8% of cited sources were at least partly AI-assisted. Crucially, the correlation between whether ai or non-ai was the cited source was effectively zero. Google doesn’t appear to be rewarding or punishing content based on its origin at all. Which, if I was in both the AI and search engine business, is probably the line I would take too.
So, it seems to mean that “write everything yourself for authenticity” instinct is wrong (caveat: it’s wrong for now) What we’re watching, in the researchers’ own words, is something close to an ouroboros: AI-generated content increasingly citing other AI-generated content, feeding a loop with less and less human involvement at any given link in the chain. That might sound dangerous, and it potentially is if we get the AI to update the operating manual for a nuclear reactor, but hallucinations aside, is probably not too much of a problem when selling washing machines.
People are noticing, even if the algorithm isn’t
Here’s where it gets genuinely interesting for anyone thinking about content strategy. While the ranking systems stay indifferent, actual consumer sentiment toward AI-generated information has been moving sharply in the opposite direction. Research from Search Engine Land and Fractl found that the proportion of consumers who considered AI-powered search more helpful than traditional search fell from 82% to 54% in a single year, even as usage of AI search tools kept climbing (Search Engine Land / Fractl, 2026). People are using these tools more and trusting them less at the same time, which is an odd but increasingly common pattern with new technology adoption.
Now, this might be an issue for Google and for the AIs that people are using for answers, but it’s unlikely to be an issue for the general population skimming your webpage. The average reading age in the UK is estimated to be somewhere between the capability of the average 9-to 11-year-old, so I doubt many people will spot the nuance behind the AI writing style.
Where brands will have issues is with AI hallucinations. Any heavy user of AI will tell you that it frequently makes things up, or only half remembers. I’m not a programmer by any measure, but I’ve vibe coded scripts for a few personal projects, and I’ve been told to use deprecated services, paid services when I’ve specifically asked for free versions, etc. So, I can definitely confirm that AI needs to be treated with the same type of care you’d apply to an imaginative 6-year-old telling you about his day at school.
For individual businesses, the research showed that 27% of brands surveyed had already been misrepresented in AI-generated responses about them, and only 24% of marketers were actively tracking how their brand appears across AI tools at all. That’s a meaningful gap, most businesses have no idea what ChatGPT or Google’s AI Overviews are currently saying about them, and roughly a quarter of those who do check have already found something wrong. The good news though, (depending on how you look at these things, is that the particular claim, or your entire citation, may disappear into the ether within minutes, in a manner that is virtually unknown in natural search.
What does this aversion to AI mean practically?
I wouldn’t read any of this as a reason to avoid AI tools in your own content production; that ship has sailed, and Ahrefs’ data suggests the algorithm isn’t punishing you for it anyway. What I would take from it is that human trust is becoming the genuinely scarce resource in a way that ranking position no longer fully captures. If the web is filling up with interchangeable, AI-assisted content that all sounds roughly the same, the content that visibly demonstrates real expertise, a specific case, a genuine opinion, a named person standing behind a claim, becomes disproportionately valuable precisely because it’s becoming rarer relative to the flood around it.
The other practical takeaway is more of an operational one. With over a quarter of brands already misrepresented by AI tools and barely a quarter of marketers checking, it’s worth building a habit of periodically asking ChatGPT, Gemini or Google’s AI Overviews what they say about your own business, the same way you’d once have periodically Googled yourself. It costs nothing, takes a few minutes, and catches a problem most businesses currently have no visibility into at all. If you are feeling particularly adventurous, set up a spreadsheet, ask it the question and catalogue the answers to see how you content varies the results over time. I can guarantee you that not many businesses, or even digital marketing people, will have picked up on that yet.
Rich Jarrott, the founder of ZenithSpark is an experienced digital marketing professional. If you have a problem, If no one else can help and if you can find them. Maybe you can hire, ZenithSpark.
Sources: