I used to use em dashes in my writing. Not everywhere, not constantly, but when they felt right. When I needed to set off extra information, emphasize a point, or expand on an idea in a way that a comma or period couldn’t quite capture.
Then ChatGPT happened.
Now I can’t use an em dash without worrying that everyone reading will assume I’m lazy, that I let AI write everything for me, that I have no idea what I’m doing.
The em dash didn’t deserve this.
What I loved about em dashes
First, a quick clarification. The hyphen (-) connects compound words. The en dash (–) shows ranges. The em dash (—) is the longest one, and it’s the one that ChatGPT has beaten to death.
An em dash is a sophisticated punctuation mark. It has real functions: setting off information, explaining, emphasizing, expanding upon what came before. It’s not decorative. It does work that other punctuation can’t quite manage.
I learned to use em dashes in academic writing. I understood when they were appropriate and when they weren’t. I used them sparingly in my personal writing because I thought online content required simpler punctuation, basic structures, maybe a semicolon at best. I was wrong about that, but I was being thoughtful about it.
The truth is, my writing could have benefited from more em dashes. They would have taken it to the next level. But I held back because I wasn’t sure they belonged in the informal, accessible style I was trying to create.
Then AI made the decision for me.
When everything changed
After ChatGPT was released, em dashes started appearing everywhere. Not used thoughtfully. Not used sparingly. Overused to the point of being unnatural, unhinged, almost an eyesore.
Suddenly everyone was writing like this: “The problem isn’t just about timing—it’s about strategy.” Or: “Success is more than just working hard—it’s about working smart.” Those structures, those em dashes, everywhere. In every LinkedIn post, every blog article, every piece of content trying to sound profound.
It became a tell. A red flag. See an em dash, assume it’s AI.
And here’s the thing: I can’t even blame people for making that assumption. The pattern is real. ChatGPT has a tendency to abuse em dashes in a way that makes them stand out. People started using them to impress, to sound sophisticated, without understanding what they actually do.
Now I read something with an em dash and my first thought is “this is AI.” Even when I have no evidence. Even when the person might have just used it as normal punctuation, the way it’s meant to be used.
But the widespread, aggressive overuse of em dashes didn’t exist before ChatGPT. That’s the difference.
The other casualties
It’s not just em dashes. AI has ruined entire sentence structures for me.
“It’s not just about X, it’s Y.”
“X is more than just [something], it’s [something else].”
These used to be perfectly fine ways to structure an idea. Now they scream ChatGPT. Now using them makes you look like you have no original voice.
I find myself avoiding these structures even when they would work perfectly for what I’m trying to say. I second-guess everything. Did I think of this phrasing myself, or did I absorb it from seeing too much AI-generated content?
The deeper problem
Here’s what really bothers me: English isn’t my first language. I have an advanced command of it, but I’m still constantly aware that it’s not my native tongue. I’m always working to make sure my writing is clear, correct, natural.
So I use AI tools to proofread. To catch errors. To help me sound more fluent.
And those tools change everything. They rewrite my sentences. They add em dashes where I didn’t put them. They restructure my ideas into those AI-typical patterns. My authentic voice gets diluted because I’m relying on something that’s actively homogenizing how everyone writes.
Then I question myself. Was my original sentence actually wrong, or just different? Am I making a mistake, or am I just writing in my own voice?
The constant second-guessing is exhausting.
A melancholic tribute
The em dash deserved better than this.
It’s a useful piece of punctuation that’s been doing its job quietly for centuries. It helped writers create rhythm, add emphasis, insert necessary asides without disrupting the flow of a sentence.
Now it’s been turned into a marker of inauthenticity. A sign that someone couldn’t be bothered to write their own content. A red flag that makes readers suspicious.
That’s not fair to the em dash. And it’s not fair to writers who actually know how to use it.
I had a relationship with this punctuation mark before AI became ubiquitous. I understood its purpose. I used it thoughtfully. Now that history doesn’t matter because everyone assumes the worst.
What I’m going to do anyway
Here’s the thing: I’m done letting AI dictate what punctuation I’m allowed to use.
I’m going to use em dashes when they feel right to me. When they serve the purpose I need them to serve. When they make my writing clearer or more emphatic or more alive.
If you read my work and assume it’s AI-generated because you see an em dash, that’s your problem, not mine. You’re making an assumption without evidence. You’re letting AI’s overuse of a punctuation mark erase the possibility that someone might actually know what they’re doing.
I refuse to let ChatGPT take away tools I understood before it existed.
The em dash didn’t ruin itself. AI ruined it. And I’m not giving it up just because other people can’t tell the difference anymore.
Featured image courtesy of Rawpixel.