Since ChatGPT was launched at the end of 2022, there’s been a huge amount of discussion about how artificial intelligence might change the way we work, and one of the areas where it’s arguably had the biggest immediate impact is digital marketing. In particular, the ability to generate entire articles, web pages and other written content in a matter of seconds has raised an interesting question – if ChatGPT can write content for us, why would we bother writing it ourselves?
It’s a reasonable question, especially when you consider how much time and money businesses spend producing content for their websites. If you can generate a 1,000-word blog post in 30 seconds rather than paying a copywriter to spend several hours writing one, the potential savings are enormous. However, there’s a rather important consideration here – what does Google think about all this, and could using AI-generated content actually damage your website’s search engine rankings?
Does Google Penalise AI-Generated Content?
The short answer is no, or at least not automatically. In February 2023, Google published some official guidance on AI-generated content, which clarified that it doesn’t necessarily matter whether content has been written by a human or generated by artificial intelligence. What matters is the quality of that content, and whether it’s genuinely useful to the people reading it.
This is an important distinction, as Google has spent the best part of two decades trying to identify and penalise websites that produce poor-quality content purely to manipulate search engine rankings. In the early days of SEO, it was relatively easy to generate hundreds of pages stuffed with keywords and watch them climb the search results. Google’s algorithms have become considerably more sophisticated since then, and the introduction of ChatGPT hasn’t changed the underlying principle that content should be written primarily for people, rather than search engines.
In fact, Google’s position is quite logical when you think about it. If an AI-generated article provides genuinely useful information that answers a user’s question, why should it be penalised simply because of how it was produced? Equally, if a human writes a terrible article that serves no useful purpose, why should Google reward it just because a person wrote it?
What About Google’s March 2024 Update?
Things became slightly more interesting in March this year, when Google introduced a major update to its search algorithms, alongside new spam policies specifically targeting what it calls ‘scaled content abuse’.
Essentially, this means producing large quantities of content with the primary intention of manipulating search rankings, rather than providing useful information. Importantly, Google made it clear that this applies regardless of whether the content is generated by AI, written by humans, or produced through a combination of the two.
You can read Google’s announcement here, but the important point is that using ChatGPT to write content isn’t inherently against Google’s guidelines. Using it to churn out hundreds of virtually identical articles targeting slightly different keywords, however, could cause problems.
This is particularly relevant to businesses that have traditionally relied on producing large volumes of SEO content. Suddenly being able to produce ten times as many articles for a fraction of the cost sounds fantastic, but if those articles don’t offer anything that isn’t already available elsewhere, what’s the benefit to the person searching?
Is ChatGPT Actually Any Good at Writing Content?
This is where things become a little more subjective. Having experimented with ChatGPT myself, I think it’s an incredibly impressive piece of technology, but that doesn’t necessarily mean it’s a particularly good replacement for a human writer.
For a start, ChatGPT has a tendency to produce content that sounds rather generic. It can explain a subject reasonably well, but it often lacks the personality, experience and opinions that make an article interesting to read. There’s also the problem of accuracy – ChatGPT can confidently produce information that is completely wrong, a phenomenon commonly referred to as ‘hallucination’.
That might not matter too much if you’re asking it to write a light-hearted article about your favourite films, but it becomes considerably more serious when you’re producing content about health, finance or other subjects where inaccurate information could have real-world consequences.
Google also places considerable emphasis on something it calls E-E-A-T, which stands for Experience, Expertise, Authoritativeness and Trustworthiness. These aren’t individual ranking factors in themselves, but they represent the characteristics Google wants its systems to recognise in high-quality content. An article written by someone with genuine first-hand experience of a subject has something that ChatGPT simply can’t reproduce independently – actual experience.
So Should We Be Using It?
Personally, I think the most sensible approach is to treat ChatGPT as a tool rather than a replacement for a writer. It’s excellent for researching subjects, suggesting article structures, identifying areas you might have overlooked and helping to overcome the occasional bout of writer’s block. It can also be useful for rewriting awkward sentences or explaining complicated concepts in simpler language.
Where I’d be more cautious is allowing it to produce entire articles with little or no human involvement, particularly if those articles are being published purely for SEO purposes. Not necessarily because Google will detect that they’re AI-generated and penalise them, but because there’s a good chance they won’t provide anything particularly original or useful.
Of course, there’s an interesting irony in all this. If Google genuinely doesn’t care how content is produced, and AI continues to improve at its current rate, we could eventually reach a point where AI-generated articles are indistinguishable from those written by humans. At that stage, does it really matter who – or what – wrote them?
For now, though, my advice would be fairly straightforward. Use ChatGPT as much as you like to assist with content creation, but make sure you understand the subject, check the facts and contribute something of your own. If the finished article is useful, accurate and genuinely worth reading, then you’ve probably got very little to worry about.
And if your entire SEO strategy consists of asking ChatGPT to generate 500 articles and uploading them straight to your website, you might want to reconsider.