Punctuation and the Dash Dilemma—
Why AI writing gives itself away with N– and M—dashes
You’re reading an article and it seems fine—polished, even. But something feels... off. There’s a strange overuse of punctuation. Sentences break in all the right (or wrong) places. And suddenly, it hits you:
This was written by AI.
What gave it away? Likely, the unnatural abundance of n-dashes (–) and em-dashes (—).
AI’s obsession with these humble punctuation marks is one of the biggest tells in machine-generated writing—and here’s why.
A Quick Dash Refresher
Let’s get technical for a moment:
Hyphen (-): Used to join words (e.g. well-known, twenty-three)
N-dash (–): Used to indicate ranges (e.g. 2010–2020) or link related concepts (e.g. London–Paris flight)
Em-dash (—): Used to break or emphasise parts of a sentence—like this.
Both the n-dash and the em-dash are valuable punctuation tools when used sparingly and intentionally. But when AI gets involved? That sparing use often goes out the window.
Why AI Overuses Dashes
It Was Trained on Perfectionism
AI is trained on large datasets full of textbooks, style guides, online encyclopedias, and professional journalism. These sources religiously follow editorial standards, using em-dashes for emphasis and n-dashes for technical accuracy. So the AI follows suit—even when it's not necessary.
It Thinks You Want "Polished"
AI is trying to impress. It wants to sound smart, fluid, and structured. In doing so, it often uses em-dashes as a shortcut for sentence variety and rhythm. The result? A whole lot of unnecessary breakpoints that start to feel unnatural.
Dashes Help With Flow—In Theory
Breaking up a sentence with dashes creates a natural pause and can avoid clunky parentheses or comma overuse. But AI, lacking a real sense of voice or rhythm, often uses them as filler, mistaking punctuation for personality.
It Doesn’t Know When Enough is Enough
Humans tend to write with feel. We know when a dash adds flair—and when it’s just annoying. AI? Not so much. It sees a grammatically correct structure and applies it repeatedly, leading to an over-dashed article that feels off, even if the grammar’s technically right.
Why It Matters
The overuse of n- and m-dashes has become a hallmark of AI-generated content, especially in blog posts, thought leadership pieces, and educational articles. For savvy readers and editors, it’s a flashing neon sign: “This was written by a machine.”
It’s not just about style—it’s about authenticity. When every third sentence is broken by a dramatic dash, readers begin to tune out. What should be a human-sounding insight starts to feel robotic and over-structured.
Human Writing Isn’t Perfect—and That’s the Point
Real writing reflects real thought. Sometimes it’s messy. Sometimes it breaks grammar rules for tone or voice. Sometimes it just uses... a plain old hyphen instead of worrying about whether it's technically supposed to be an n-dash.
In contrast, AI writing can feel like a student desperate for an A+ on every paragraph—overcorrected, too safe, and just a little too eager.
The Bottom Line
If you're using AI to help write content (which is perfectly fine), watch out for the dash addiction. Edit with a human eye. Swap out overused dashes for simpler punctuation. Ask whether the rhythm of the sentence really calls for a dramatic pause—or if a full stop might serve better.
Because in the end, it’s not the grammar that builds trust—it’s the voice. And ironically, the clearest sign that something wasn’t written by a human... is perfect punctuation.
This blog was constructed using AI.
Published 29th March 2025
