Top 10 Tells That Someone’s Using ChatGPT

The question among fundraising copywriters is no longer, “Do you use ChatGPT?” but “How do you use ChatGPT?” The artificial intelligence (AI) tool writers feared would take their jobs has quickly become too useful to ignore. It’s a classic case of “if you can’t beat ‘em, join ‘em.”  

Fundraisers, in particular, should care — because if your appeals, thank-yous or campaign updates sound robotic, donors will notice it immediately. That doesn’t mean it’s not still coming for your job, though the consensus is to at least get something out of it while it’s still useful. 

One challenge is that, like AI-generated images, AI-written copy is becoming easier to spot. So, here’s a very abbreviated guide to help writers who use it save face — and stay honest — by recognizing 10 of ChatGPT’s most overused rhetorical devices. As more people become familiar with them, the chances are greater that you’ll get busted if you over-rely on its bad habits like: 

1. Hackneyed Transitional Phrases

ChatGPT outputs lean heavily on transitions like:

• “That being said”
• “All in all”
• “In other words”
• “At the end of the day”

As a dedicated copywriter, you should avoid these trite phrases anyway, but now, if you rely on them too much, you’ll look even more suspect.

2. Synthetic Parallelism

This is the device in which one phrase establishes a concept or premise, and the next phrase clarifies or contrasts it. 

• “It’s not just a job, it’s a career.” 
• “You’re not just helping, you’re changing lives.”
• “It’s about more than winning — it’s about sportsmanship.” 

You get the idea. These kinds of constructions have always been a little overbaked anyway, so this is an opportunity to cut out the dramatics and just let language do its job, e.g. “You’re changing lives,” or “It’s about sportsmanship,” etc. 

3. Predictable Introductions

Most ChatGPT articles begin with unoriginal lines like: 

• “In today’s world”
• “As technology continues to evolve”
• “Whether you’re a beginner or an expert”

If an opening sentence sounds like it was pulled from a stock template, it just might be ChatGPT.

4. Excessive Balance

Like an old-fashioned politician, ChatGPT often evenly balanced responses, defaulting to evenhandedness. Unless you push it to be opinionated in your prompt, it defaults toward trying to show both sides of an issue — even when a strong opinion makes more sense.

5. Short on First-Person References

Obviously, ChatGPT has never had any actual experiences — and it shows. Authentic first-person storytelling builds connection. To get realistic-sounding first-person copy, you must direct it carefully and hope it has culled enough examples from the internet not to embarrass you. Its lack of actual experience also makes it avoid specifics like names, places, dates, details and especially emotions. 

6. Redundant Structure

Listicle language that helps you organize your thoughts can make life easier for both you and your reader. However, ChatGPT doesn’t seem to know when enough is enough. You might be dealing with someone who’s grown a little too dependent on ChatGPT if you feel battered and bruised by too many paragraphs beginning with clunky phrases like: 

• “First, we’ll discuss…”
• “Second, let’s examine…”
• “Third, we’ll consider…”

7. Lack of Sentence Variety

Good writers vary sentence length to keep their thoughts on track, provide the exact amount of information they want you to have in one thought and even create moods. (If you like writers like Melville or Henry James, you’ve seen extreme examples of that.) 

Modern writers don’t force you to wade through paragraph-long sentences anymore, but neither do they hesitate to use two-word sentences next to ones that take up three or four lines to be expressive. ChatGPT-generated text tends to avoid those contrasts. 

8. Overuse of Select Transition Words and Phrases

ChatGPT outputs frequently include select transition words and phrases that sound formal and summarizing. For instance:

• “Ultimately” is its go-to word in conclusions.
• “Indeed” is overused in emphasizing points of argument.
• “In essence” is the most annoying because ChatGPT uses it to summarize points it has already summarized.

9. Pointless Filler Copy

People often give ChatGPT a word length to adhere to. To be obedient, it will load up its writing with meaningless, throat-clearing phrases like:  

• “It goes without saying that”
• “In this day and age”
• “As previously stated”

Again, many people write this way, but ChatGPT doesn’t know when to stop.

10. Trite Metaphors

To the point that even the most mediocre writer would blush, ChatGPT outputs overuse clichés and lame metaphors, such as:

• “Double-edged sword”
• “Tip of the iceberg”
• “Slippery slope”

It can’t invent creative metaphors unless it’s creatively prompted.

The bad news — and the good news — is that many of these devices and expressions are widely used even without AI. So, this list can also be a challenge to up our own games and find more creative ways to express ourselves. In the end, that can only be good for the writing trade.

Two short postscripts: First, all the above points are true as of late summer 2025. But ChatGPT learns at an astonishing rate, so who knows how sophisticated it will be in the months and years ahead. 

And second — did I use ChatGPT to help me write this article? Of course I did. But if I’ve done my job well, you’ll have some trouble picking out where and how often. 

The preceding content was provided by a contributor unaffiliated with NonProfit PRO. The views expressed within may not directly reflect the thoughts or opinions of the staff of NonProfit PRO.