In 2025, humans rely on AI more than ever. While it can help us in numerous ways, its most common use is in the form of a chatbot – ChatGPT is undoubtedly the most popular one. So, how to tell if something was written by ChatGPT? It’s a question that leaves many wondering. Can you tell if someone used ChatGPT at all? Chat might be good, but it’s not undetectable. Let’s go through a few tactics for spotting AI-generated content.
How to Tell if Something Was Written by ChatGPT?
How to tell if ChatGPT wrote something? When you ask somebody this question, nine out of ten times, the answer will be – use an AI detection tool. Of course, this is a valid answer, but it’s not the only one, and it’s certainly not the best one. Any experienced marketer who understands content creation can tell you that these AI detection tools are not always accurate. They don’t get it right every time, and sometimes, AI-written content just slips through the cracks. If you want to know how to spot ChatGPT effortlessly, you must learn the tricks this AI uses.
Fortunately – or unfortunately, depending on how you look at it – Chat is still not good enough to fool humans completely. It will inevitably become better in the next few years as OpenAI, the company that created it, puts a lot of effort into its improvement. But for now, we can still tell AI-generated content apart from what a content writer has written. Let’s learn how to check if something was written by ChatGPT – there are a few obvious telltale signs.
Chat GPT Uses a Few Common Phrases All the Time in Its Text
When you hire a digital marketing agency and a content writer to create a blog post or some other type of content for you, the product you get is a creative, original piece that will captivate the reader’s attention and make them want to read till the end. The value of content writers is in their creativity and imagination – they can come up with phrases and sentence constructions that artificial intelligence just can’t match. ChatGPT can write various types of content, but it simply won’t be as good as human-written text.
When we ask ourselves how to tell if something is written by ChatGPT, the first clue to look for is the usage of several phrases that (for some reason) Chat loves. These include unnecessary constructions like in summary, it’s important to note, a testament to, delve into, a rich tapestry, and in conclusion. Another popular one is the phrase ever-evolving landscape – we can all agree it’s not a phrase that writers use every day, right?
While writers will sometimes add these word constructions to their texts, ChatGPT uses these phases excessively. They add a pretentious tone to the text that you will rarely see in human-written content. It almost looks like ChatGPT is trying too hard to sound smart, but in reality, it’s giving itself away. The more you read AI-generated content, the easier it becomes to spot repetitive phrases Chat uses.
The Rhythm of the Text Is Predictable and Monotonous
It’s not just about the words – it’s also about the way they’re used. When a person writes, they do their best to avoid repetition and monotony – unless their goal is to make the reader fall asleep. On the other hand, ChatGPT loves patterns, so it writes in a predictable way. Its sentences are similar, their structures rarely vary, and the length of paragraphs is always the same. If a content writer tried to write so monotonously, they wouldn’t be able to – writers naturally try to create writing with a flow that will make it easy to stay focused while reading.
Chat Uses Bullet Points a Lot
Another great answer to the question “How to know if ChatGPT was used?” is this – look for excessive use of bullet points. Now, you might ask – isn’t using bullet points good for the structure of your blog post? It is, but in this case, there can be too much of a good thing. Chat’s bullet points are often just listed words without any explanation, or it can write points where commentary is redundant and doesn’t provide additional information about the topic in question. In most cases, it’s quite easy to spot a bullet point that’s AI-generated.
The Content Generated by AI Is Rarely Specific
The next point on our list of tactics on how to tell if ChatGPT was used is a bit harder to learn – it’s a bit less obvious than the telltale signs we’ve mentioned above. Chat rarely writes ultra-specific texts that explore the topic in detail from various angles (or, as Chat would say, it doesn’t delve into the topic quite successfully). You can spot this, but only if you know a bit about the topic you’re reading on – if you don’t, it can be hard to distinguish basic information from useful insights.
However, one thing you can count on when it comes to AI writing is that it won’t offer you actual, useful advice. It will say something as far from specific as possible – an obvious advice that you could’ve thought of by yourself. ChatGPT is like a friend that tells you not to be sad if you tell them you’re sad or to just go to sleep if you say you’re tired. On the other hand, if the text you’re reading is written by an actual human expert on the topic, you’ll feel like it provides useful suggestions.
Chat Can Often Include Misinformation in Its Texts
Arguably, the biggest problem with chatbots like ChatGPT is the fact that they commonly provide false information. If you’re writing something with Chat, you want to double-check everything important, as the chatbot can easily make up false claims or misquote sources. It can quite literally make up information that has no basis in reality – this is known as AI hallucinations.
AI hallucinations occur as a result of Chat’s nature. This platform is designed to predict text based on available public information and training datasets. In other words, Chat is trained to present people’s assumptions as facts. We must keep in mind that Chat is powered by narrow AI. This means it’s created for a specific task – it can detect patterns and correlations in datasets to make predictions, but these datasets must be of high quality and accurate. The reason for misinformation written by ChatGPT lies in the datasets it uses.
Chat’s Content Is Generic – It Lacks Any Personality
Artificial intelligence doesn’t have its own personality like a human does. Writers always implement some aspects of their personality into the texts they write, whether they notice it or not. You can try to write something 100% impersonal, but unless you’re writing some technical manual, the chances are that you’ll fail at that task.
On the other hand, ChatGPT writes impersonal texts ten out of ten times – it just doesn’t know any other way to write. You can easily notice that AI-generated content lacks personal insights, jokes, and anecdotes. It might contain some metaphors, but they are usually strange metaphors that humans would never use.
It’s Easy to Spot That AI-Generated Texts Lack Emotion
Just like it doesn’t have a personality, AI also lacks emotion. It might be great at many things, but it still hasn’t reached the level at which it can mimic human emotions (and do we even want it to be able to do that?!). The texts written by ChatGPT feel emotionless or have a tone that doesn’t fit the topic. For example, the text is supposed to be about something sad, but it has a robotic tone – no writer would be able to write about a sad topic without acknowledging that it’s sad or writing something comforting.
The Longer the Text, the Easier It Will Be for You to Realize It’s Written by AI
If you ask ChatGPT for a short text about a certain topic, the chances are that it will be able to provide you with decent content. However, if you request a longer text on the same topic, you’ll get something that everyone could spot as AI-generated from a mile away. Why is this the case? When it writes a short text, Chat can rely on its pre-trained patterns.
However, as the text goes on, the mistakes start to occur. Chat starts to repeat the patterns it uses. You can see the repetitiveness in phrases, sentence structures, vocabulary, and even information. It will look like Chat is starting to blabber more and more as the text length increases until it stops making sense. So, when you want to spot AI-generated content, we suggest you read more than the first few paragraphs. If the text is lengthy, its quality will decrease as you read on.
Why Is It Always Better to Post Human-Written Content?
Now that you know how to tell if someone used ChatGPT, the question is – why is this important for anyone who has a website (and wants to get more people to that website)? As we can all agree, the content written by ChatGPT is not exactly the best – content writers can do it so much better. If your website is there to represent your small business in the digital world, you likely want to fill it with quality content and useful, meaningful text.
But that’s not the only reason why you should always choose human-written over AI-generated content. Search engines like Google know how to spot text written by ChatGPT, and let’s put it this way – Google is not a fan of AI-written blog posts. It will rank worse than original, human-written content. Why?
It lacks originality and value and doesn’t provide any new information. If you understand the basics of SEO, it will be clear to you why AI-generated content ranks badly and why reputable businesses would always choose content writers over ChatGPT.
Hire a Reputable Content Writing Agency and Ensure Your Website Is Filled With Valuable Content
You’ve decided to hire a content writer to create excellent blog posts for your website – now, the question is where to find a good writer. The answer is simple. All you need to do is hire a content strategy agency that can provide you with content writing and SEO services – like our Made Online. We are a digital content agency that specializes in content writing. Our team consists of several experienced content writers who can effortlessly create outstanding blogs on any topic. Contact us today to schedule an audit with our team, and let’s figure out what your website needs!