With the rise of AI, and business owners and marketing managers becoming more educated about SEO, we’ve seen an increase in clients wanting to take parts of their content and SEO strategy into their own hands. For the most part, I think this is a good thing. Clients are becoming more educated, they’re interested in how Google works, they understand that content plays an important role in search rankings, and they’re embracing new technology rather than ignoring it.
Where we’re becoming concerned is how some businesses are actually using AI. We’re hearing about businesses wanting to generate hundreds, sometimes thousands, of articles targeting every imaginable variation of a search term. We’re also seeing AI SEO plugins that can do the keyword research, come up with the article idea, write the article, create it inside WordPress and publish it automatically.
On the surface, that sounds incredible. You can suddenly do in a week what might previously have taken a marketing team a year. But there’s a pretty big problem: Google doesn’t reward you simply because your website has more pages.
In fact, in its latest guidance around AI and Search, Google has specifically told website owners to focus on creating valuable, unique, non-commodity content and warned against creating separate pages for every possible variation of what somebody might search. Google explicitly says that having a high quantity of pages does not make a website higher quality or more relevant.
I think a lot of businesses are about to learn this the hard way.
The thinking that more content equals better SEO
Content absolutely plays an important role in SEO. At a very basic level, it allows you to target the topics and search terms relevant to your business. More importantly, it gives you an opportunity to demonstrate experience and expertise around those topics.
The mistake is turning that into the idea that if 10 articles are good, 100 must be better.
We’ve had conversations with clients who are looking at producing huge numbers of AI-generated blog posts, location pages and variations of service pages because they believe every additional page creates another opportunity to rank. Technically, every page is another URL Google could index, but that doesn’t mean Google will index it, and it definitely doesn’t mean Google will rank it.
Google’s own Search Essentials make this clear. Meeting the technical requirements for Google Search does not guarantee that Google will crawl, index or serve a page in its results.
This is where the difference between creating content and creating useful content becomes incredibly important.
Is AI content bad for SEO?
No, and I think this is where the conversation about AI content often gets oversimplified.
Google isn’t saying that using ChatGPT automatically makes an article bad. Google specifically acknowledges that generative AI can be useful for researching topics and helping structure original content. The issue is using generative AI to produce large numbers of pages without adding value for users. Google says doing that may violate its policy against what it calls scaled content abuse.
That’s an important distinction. Using AI to help produce a genuinely useful article isn’t the problem. Using AI to generate 500 articles because you’ve identified 500 keywords you want to rank for is something very different.
When the primary purpose becomes manipulating search rankings rather than helping the person reading the page, you’re moving directly into the territory Google’s spam policies are designed to target.
The 2026 version of an old SEO shortcut
We’ve been around websites long enough to see plenty of SEO strategies come and go. One that immediately comes to mind is link farming and buying huge numbers of low-quality backlinks.
This became particularly prominent throughout the late 2000s and early 2010s. In April 2012, Google released the webspam algorithm that became known as the Penguin update, specifically targeting tactics including keyword stuffing and manipulative linking schemes. Google said at the time that the update affected around 3.1% of English search queries.
At the time, the logic wasn’t that different from what we’re seeing now. If backlinks help rankings, let’s create thousands of them. Today it’s ‘if content helps rankings, let’s create thousands of pages’.
I’m not saying AI-generated content is inherently black hat SEO. It isn’t. But using AI to mass-produce low-value pages primarily to manipulate rankings is, in my view, the 2026 version of the same mentality. You’re taking something that genuinely helps SEO and trying to scale it to the point where the value behind it disappears.
And Google has seen this story before.
We’re seeing more pages being crawled but not indexed
Something else we’ve been noticing across websites is an increase in URLs sitting inside Google Search Console under Crawled – currently not indexed.
This is an important distinction. Google has found the URL and Googlebot has visited the page, but Google has decided, at least for now, not to include that page in its index.
There can be plenty of reasons for this. Technical issues can contribute. Server availability, crawling problems, poor internal linking, duplicate or canonicalisation issues, robots directives and website architecture can all affect how Google discovers, crawls and indexes content.
Website performance can also affect Google’s ability to crawl a large website efficiently. Google says slow server responses and rendering can reduce crawling efficiency, particularly where server resources are limited.
But there’s another line in Google’s documentation that I think website owners should pay much more attention to. Google says pages may not appear in search results even after they’ve been crawled if there isn’t sufficient value or user demand for the content.
Google also lists low-quality content as one of the possible reasons a page may not make it through the indexing process.
That doesn’t mean every URL showing as “Crawled – currently not indexed” is AI-generated or low quality. It would be wrong to make that assumption. But when we look at a site that has suddenly generated hundreds of very similar articles, and a large percentage of those URLs are being crawled but not indexed, you have to start asking a bigger question:
Does this content actually add anything to the internet?
Getting Google to find a page is only half the battle now. You also need to give Google a reason to keep it.
I actually think this is the biggest content opportunity we’ve ever had
This might sound strange after everything I’ve just written, but I think AI has created one of the biggest content opportunities businesses have ever had. There is more generic content being created than ever before, which makes genuinely useful expert content more valuable.
Customer expectations have also changed dramatically. People want extremely specific questions answered instantly, and who is better placed to answer those questions than the people actually doing the work every day?
Take a mechanic. Imagine you’re thinking about buying a new car. You’ve narrowed it down to three models, watched the reviews, taken one for a test drive and everything looks great. Eventually you’re probably going to ask: is this thing reliable?
You can read manufacturer information, read car reviews or ask ChatGPT. But I’d want to talk to the mechanic who has worked on thousands of cars. What brands are constantly coming back into the workshop? What fails at 100,000 kilometres? Which cars are expensive to repair? Which models rarely come through the door? What would the mechanic buy with their own money?
Now imagine that mechanic writes an article called:
The Cars We’d Buy in 2026 After Working on Thousands of Them
Inside the article they explain the common problems they see, which engines they trust, which models they would avoid, what repairs typically cost and which vehicles they hardly ever see because they simply don’t break.
That’s valuable content. More importantly, it’s information ChatGPT couldn’t genuinely produce on its own. AI can help the mechanic turn those thoughts into a structured article, but it can’t replace the experience that makes the article valuable.
That’s the opportunity.
A real example from our own website
We recently experienced this ourselves when WordPress 7.1.1 was being released and we published an article about the update.
We could have simply asked ChatGPT to write a blog post about WordPress 7.1.1, and it probably would have produced something perfectly readable. But that wasn’t what we did.
We looked at how we were actually approaching the update as a web agency. How were we rolling it out across the websites we manage? Had our WordPress update process changed? Which websites could be updated automatically? Which ones needed manual testing? How did we approach more complicated websites with custom functionality? What did this update mean for the ecosystem we manage every day?
Those ideas came from us. AI helped turn them into an article, but the substance wasn’t generated by AI. The experience, opinions and process behind it came from running an agency and managing WordPress websites every day.
The article ended up reaching the top few results in Google for searches around “WordPress 7.1.1” and produced a significant spike in traffic to the website when interest around the update peaked.
We also saw improvements across a number of other search terms around the same period. I wouldn’t claim that one article single-handedly caused every ranking improvement across the website because SEO doesn’t work that cleanly, but it’s a great example of the type of content we want to produce more of.
It answered a timely question and added something that wasn’t available by simply summarising the WordPress release notes. It gave our perspective.
What should businesses be publishing instead?
Our recommendation is actually to use AI. We use it ourselves.
The difference is that we use it as a production tool, not as the source of our expertise.
Before publishing something, I’d work through roughly this process:
- Start with a real question. What are your customers actually asking you? What decisions are they struggling with? What information would genuinely help them?
- Get the expert involved before AI writes anything. Interview the person who does the work and get their opinions, experiences, examples, disagreements and recommendations.
- Add information AI couldn’t know. Real pricing, project examples, screenshots, processes, mistakes you’ve seen, customer outcomes, original photos, internal data, failures and lessons learned.
- Use AI to organise the thinking. This is where AI is extremely good. Use it to structure the article, identify gaps, improve readability, research supporting information and turn rough thoughts into something people actually want to read.
- Fact-check it. Particularly if you’re publishing advice around technical, financial, legal, medical or rapidly changing topics.
- Ask why this page deserves to exist. If your article basically says the same thing as the first ten Google results, go back and add something original.
- Publish because the article is useful, not because your content calendar says you need another blog post this week.
This is essentially how we’re approaching our own content. AI allows us to produce articles more efficiently, but it doesn’t replace the expert behind them.
Google’s message is actually pretty simple
There are hundreds of SEO ranking factors, constant Google updates and an entire industry dedicated to trying to understand exactly how Google’s algorithms work. But Google’s latest advice around AI content is surprisingly simple: create content your visitors will actually find useful.
Google’s newer guidance for generative AI search specifically recommends valuable, unique, non-commodity content. It also warns businesses not to create pages for every possible variation of a search query simply to manipulate rankings.
To me, that’s the most important part of this entire conversation. The question isn’t: “Was this written by AI?” The better question is: “Would this content exist if we weren’t trying to rank for a keyword?”
If the only reason an article exists is because an SEO plugin found a keyword and automatically generated 1,500 words about it, I’d be questioning whether it needs to exist at all.
But if the article answers something your customers genuinely want to know and contains experience, insight or information they couldn’t easily find somewhere else, then AI can be an incredible tool for helping you produce it.
We’re not moving into a world where content matters less. I think we’re moving into a world where generic content matters less and genuine expertise matters more. And for businesses that have spent years building that expertise, that’s a massive opportunity.
