No, AI content isn't bad for SEO when you use it intelligently and as part of a process, where you do the research, fold in your own opinion and check every fact before it goes live. It starts to hurt when you just ask ChatGPT to write a blog on a topic and publish whatever it gives you back.
If you've started using AI for your blogs or product copy, or you've been asked to by someone, then the chances are you're either worried Google will spot it and push your site down or quietly hoping it'll save you a few hours a week. You're certainly not alone, as the ONS found that large language models were the most widely used type of AI among UK businesses with 10 or more employees in June 2026. I use it for my own writing too and I think it can be great, but good, unique content that nobody else has created takes effort and there are no shortcuts to it.
The Trouble With "Write Me a Blog On…"
A client of mine in the medical sector had been writing their blogs by just asking ChatGPT for them, which I completely understood as they were busy. You could tell they'd typed something along the lines of "write a blog on blood pressure" though, because what came back was thin and basic with nothing in it that was unique to them or of much value to the people reading it, so I had to talk them through why that was a problem.
The trouble with that prompt is that there's nothing of yours in it. All the tool can do is draw on what's already been written about blood pressure, so you end up with a tidy summary of the same points every other health site covers, and any competitor with the same tool can ask the same question and get more or less the same page back. What you don't get is any first-hand experience or anything about how you actually do things, and those are generally the bits that make someone choose your post over everyone else's.
You have to be especially careful with health, money and safety topics, which SEOs call YMYL (your money or your life). Google's guidance on creating helpful content says its systems give "even more weight to content that aligns with strong E-E-A-T" (experience, expertise, authoritativeness and trustworthiness) where a topic could significantly affect people's health, financial stability or safety, and it asks whether your content shows "first-hand expertise and a depth of knowledge". A generic AI blog on a health condition doesn't show much of either and readers tend to notice too, so a shortcut like this can damage your brand with your customers as well as with Google.
What Google Actually Says About AI Content
Generally speaking, Google's rules only really start to bite when lots of content is churned out with nothing useful added. Its spam policies define scaled content abuse as "when many pages are generated for the primary purpose of manipulating search rankings and not helping users", and they go on to say it's usually about large amounts of unoriginal content that adds little or nothing for the reader, "no matter how it's created". The guidance on using generative AI content says much the same from the AI side, that using AI tools "to generate many pages without adding value for users may violate Google's spam policy on scaled content abuse".
So no, Google won't penalise a page just because AI helped write it. Sites that break the spam policies "may rank lower in results or not appear in results at all", so the risk is real, but a single post you've drafted with AI, then checked and reworked because you know the subject, is a long way from that.
I think the March 2024 core update is a good example of where Google's attention is, as that's when it tightened its spam policies to go after content made at scale to boost rankings, "whether automation, humans or a combination are involved". Once the rollout had finished it said people were seeing 45% less low-quality, unoriginal content in search results, which was more than the 40% it had expected.
As for labelling posts as AI-written, the helpful content guidance only suggests telling readers how automation was used "when it would be reasonably expected", and the one firm rule is for AI-generated product titles and descriptions in a Merchant Center feed, which has nothing to do with organic rankings.
What the Studies Show (Take Them With a Pinch of Salt)
There have been a couple of big studies on how AI content ranks and both come from SEO tool vendors, so I'd treat them as interesting more than anything. Ahrefs looked at 600,000 pages ranking for 100,000 keywords and found only 4.6% were entirely AI-generated while 81.9% mixed AI and human input, with no clear relationship between how much AI a page used and where it ranked. The Semrush study looked at 42,000 blog pages ranking for 20,000 keywords and found pages in position one had an 80.5% probability of being human-written against 10% for AI-generated, but it sorted the pages with an AI detector and those tools are far from an exact science. What I take from both is that most of what ranks already has some AI in it somewhere, so it's how you use it that counts.
How I Actually Use AI for Content
I use AI to help with research, drafting and polishing, and the best way I can describe it is like talking to a ghostwriter. The ideas and thoughts are all mine and it just does the typing for me, and I have final editorial review.
When I'm researching a piece it helps me find the data I need to make sure any claims stand up, and I like to go back to the original source for anything I quote as these tools can get things wrong and still sound completely sure of themselves. Rather than sitting there with writer's block I can talk to it, tell it what I want to create and steer it as I go. Then once I'm happy with the piece I've created it can give it a once-over to check for any holes, areas for improvement or bits that might not be clear, and I decide what's worth changing.
For your own content, I'd start with what you can bring that the tool doesn't have, such as your own opinion on the topic, the questions your customers actually ask you or data from your own business. I'd treat every fact, figure and date it gives you as unconfirmed until you've checked it against its source, and I'd only publish as much as someone has time to properly read and edit, as a pile of unchecked posts tends to cause problems further down the line.
If you've already published a lot of AI content that nobody has really checked, I'd stop and take stock before adding any more. An SEO audit is a sensible place to start as it'll show you which posts are doing their job and which ones need reworking with your own knowledge or folding into a stronger page.
Whether you're working out where AI fits into your content or you'd just like a second pair of eyes on what's already live, have a look at my SEO services or drop me a message and we can have a chat about it.
Frequently Asked Questions
Does Google Penalise AI Content?
No, Google doesn't penalise content just because AI helped write it. Its spam policies go after scaled content abuse, which is lots of pages made mainly to manipulate rankings, whether a person or a tool produced them. If you've researched a post properly, checked it and added your own knowledge then it's fine.
Can SEO Be Done With AI?
Parts of it can, and I use AI for research, drafting and polishing my own content. You still need someone who knows SEO and your business to steer it, make the editorial calls and check every claim against its source, though, as the tool has no idea what your customers need or what makes you different.
Is SEO Still Worth It With AI?
Yes, and I'd say most of what gets you seen in AI search is just good SEO. People are still searching for what they need but they're doing it in a different way, and Google says there are no additional requirements to appear in AI Overviews or AI Mode, so the basics of a helpful, well-optimised site still apply.
