9 tips to find out if someone is using AI in content

Well, AI doesn't have to be wrong in the basics at all. You can certainly write a good piece of text with ChatGPT, for example.

But of course, the idea is not to just copy and paste a piece of text after getting it with a simple prompt.

Those lyrics are usually vacuous, not always true and often miss the mark.

The same goes for social media content. I myself am fairly fanatical on LinkedIn and see this bland AI content popping up there more and more often.

Your gut feeling then already tells you that something is not right. That's probably why you landed on this article and want to know how to recognize AI content. With these tips, you can confirm whether your gut feeling is correct.

Contentless terms

You probably recognize them.

'Welcome to the wonderful world of ...'

'The Future of ...'

"Experience the power of...

'Step inside...'

All that kind of vacuous and commercial crap ChatGPT often uses. So we have excluded these kinds of terms when using ChatGPT.

These are really not terms that a good copywriter uses every paragraph. So if you see them a lot, you can guess.

Writing Style

When ChatGPT writes a text, it is often neutral text with flat information. There is virtually no emotion in it.

In addition, sentence structures can also lack logic and sometimes there is a lack of coherence between sentences or paragraphs.

It then feels a bit like it's stuck together, without thinking about it.

Add to that the that punctuation doesn't always run smoothly either and you know for sure: this text was written with AI.

Content that is not correct

ChatGPT or other AI tools sometimes say things that are not true.

Sh*t in = sh*t out.

That can be simple things like a year, but it can also be something like the inventor of something. If it doesn't make sense, that could be a sign of AI.

The author

Is the author of an article unknown or nonexistent? Really can't figure out who is behind the article?

Or do you see a big difference in articles written a year ago and one written recently?

If so, that may also signal that AI wrote the content.

A sharp increase in number of articles written per month may also be a sign of AI-generated content.

Thinking logically

This is really an open door though. But you read a text and think: here is not right.

Something doesn't make sense, things don't add up or you feel like a person won't write like that.

You can't put your finger on it exactly, but your gut feeling tells you that something is wrong. If so, that could mean that the text was written entirely with AI.

Check for opinions or other forms of unique content

I said it before: ChatGPT often writes neutrally, emotionlessly and it is not creative and unique at all.

This also means that there are often no opinions or unique perspectives in a ChatGPT text, even though people often give them. Also unintentionally.

Another red flag when it comes to content created with AI.

Upper Case Only In Titles or in the Post

This Is The Method For Greater Success In Your Online Marketing.

Or:

This You Need to Know When Getting Started with AI.

I always read it up in my head as the next Spongebob meme:

And so I no longer take the reader seriously at all.

Those capital letters. That is also a sign that someone used AI.

Too many emojis

Just as capital letters give it away, the amount of emojis often gives it away as well. ️

Especially when emojis are used with capital letters together. Then you really catch on immediately.

So if you see one after practically every sentence, then you know that our friend ChatGPT has also been used.

AI detection tools

In addition to all the aforementioned points, you can also use a tool where you can throw in the text.

The tool then looks at all sorts of facets (including things I mentioned in this article) and then tells you if there is a chance that the text was written with AI.

A few examples of these tools:

• Originality.ai
• Winston AI
• Turnitin
• GPTZero
• Copyleaks

These tools are pretty good at what they do. And often the results are correct. But I would always use an AI detection tool in conjunction with the other tips.