Dash Free Zone

I asked AI to drop the em dash nearly 60 times. One line of code did what the asking couldn't. Then I had 25 of them.

The short answer

If you want AI to stop writing em dashes, asking helps. In my experience, it never gets you all the way to zero.

What got me to zero was a check at the very last step, right before the text goes out.

If you build with AI, that check is a line of code. If you write in ChatGPT or Claude, that check is you, and it takes about ten seconds. Both are in the steps at the bottom.

Three marks come up a lot here. The em dash is the long one, about as wide as a capital M. The en dash is the middle one you see in ranges of numbers. The hyphen is the short one on your keyboard, the one that holds a word like set-aside together.

Everything below happened while I was writing with AI every day, mostly Claude.

I used to like them

I liked em dashes. I'm a designer, and I learned the shortcut for one over a decade ago, when I got into typography. They looked good on the page. A little bougie, so I only used one now and then.

Then AI started putting them in everything, and they turned into a giveaway. I stopped using them and added them to my ban list.

ChatGPT never really stopped for me. I asked over and over. Even after I trained it pretty well, they kept showing up.

The em dash is probably part of why I switched to Claude. It got much better at it, like an annoying habit you can actually break. It still slipped sometimes, though. That's why the rule ended up in code.

I wrote about spotting them in The Em Dash Epidemic in November 2025. I'd started noticing AI writing everywhere in 2023, and the dashes were how I learned to see it. This is the next part: keeping them out.

It wasn't just me seeing things. A 2026 preprint, not yet peer reviewed, looked at about 70,000 medical research papers that had been posted before review. Before ChatGPT, about 4 in 100 had an em dash in the discussion section. Afterward it was closer to 12 in 100, and in 2025 alone, about 20.

The study is careful about what that means. It shows a shift across a whole field of writing. It can't tell you whether one paper was written with AI.

If a study that size can't judge one paper, one dash in your post can't either.

Nearly 60 times

My chat history with Claude goes back to early June 2025. By the middle of that month, the rule is already sitting in my prompts.

Nine days later I asked for it to be built in somehow, so it would stick.

It didn't stick. I went back through my chats and read every short message where I mentioned dashes. Nearly 60 of them tell it to drop the em dash, in 54 different chats. Most landed in October and November 2025. The last one is from May 2026.

From my own chat record Times I told AI to drop the em dash, by month, June 2025 to November 2026 Short messages of mine that tell Claude to drop the em dash, read one by one: 59 in all, 19 in October 2025 and 22 in November 2025, then one or two a month, the last in May 2026. 0 5 10 15 20 25 Jun 2025: 3 corrections, out of 682 messages of mine on record that month J Jul 2025: 1 correction, out of 673 messages of mine on record that month J Aug 2025: 1 correction, out of 654 messages of mine on record that month A Sep 2025: 2 corrections, out of 4,256 messages of mine on record that month S Oct 2025: 19 corrections, out of 8,757 messages of mine on record that month O Nov 2025: 22 corrections, out of 7,873 messages of mine on record that month N Dec 2025: 7 corrections, out of 7,660 messages of mine on record that month D Jan 2026: 0 corrections, out of 3,427 messages of mine on record that month J Feb 2026: 1 correction, out of 2,414 messages of mine on record that month F Mar 2026: 0 corrections, out of 802 messages of mine on record that month M Apr 2026: 1 correction, out of 1,336 messages of mine on record that month A May 2026: 2 corrections, out of 1,736 messages of mine on record that month M Jun 2026: 0 corrections, out of 757 messages of mine on record that month J Jul 2026: 0 corrections, out of 414 messages of mine on record that month J Aug 2026: 0 corrections, out of 647 messages of mine on record that month A Sep 2026: 0 corrections, out of 758 messages of mine on record that month S Oct 2026: 0 corrections, out of 2,178 messages of mine on record that month O Nov 2026: 0 corrections, out of 817 messages of mine on record that month N 19 22 2025 2026 Times I told AI to drop the em dash, by month
Short messages where I told Claude to drop the em dash, read one by one. 59 in all.

My record has gaps, so that's a floor. It also gets much thinner in 2026. So the drop doesn't prove it learned. It could be better models, or just fewer chats on record.

My ChatGPT master prompt had the same rule by September 2025.

I wasn't the only one stuck on this. On November 13, 2025, the same day I published the epidemic piece, OpenAI's CEO posted that if you tell ChatGPT not to use em dashes in your custom instructions, "it finally does what it's supposed to do."

An engineer who wrote about it put it simply. It's surprisingly hard to prompt models out of em dashes, and nobody really knows why they love them so much.

It swapped one tell for another

When it did listen, it found a way around me.

In October 2025 hyphens started showing up where the em dashes used to be. A hyphen with a space on each side, in the middle of a sentence, is a stand-in for the dash. I had to ban those too.

In May 2026 I had a script clean up my framework library. It went through 150 frameworks and made 3,937 swaps. 2,551 became commas. 1,343 became colons.

So a third of them turned into colons.

This part is my read, not a measurement. Replace one mark with one other mark everywhere, and the tell moves with it. A page full of colons where the dashes used to be is its own pattern.

I watched the same thing happen with images that August. Telling a model to make a picture look less like AI just traded one even pattern for another. What worked was pulling rules back out and looking at the result.

The day the prompt lost

In May 2026 I built an Android app that runs one of my frameworks. It talks to Claude through a small server I control, a middle step between the app and the AI.

One of its first real answers came back with an em dash in it.

So we added a strong rule to the instructions that server sends with every request: never use em dashes.

The next answer still had one, right between a score and its label.

So we put a replace in the server's code. Every em dash gets swapped out before the text reaches the app, on every screen. That fixed it completely.

Ars Technica explained why, writing about that post. A normal program follows a rule like "don't include character X" exactly. For a language model, the same request shifts the odds. It isn't a hard rule. Its own quick test found the custom instruction did seem to work.

So the line in the prompt lowers the odds. The line in code is the guarantee. My own writing rules say that now, in so many words: enforce it in code where the text comes out.

The chat assistant on my sites carries both.

From the live site chatbot
// In the prompt: it asks
STYLE: Never use the long em dash punctuation in your replies. Use a period or a comma instead. This is a firm brand rule.

// In the code: it makes sure
$response = preg_replace('/\s*(?:\x{2014}|\\\\u2014)\s*/u', ', ', $response) ?? $response;
Both lines are in the chatbot on my sites today. The second runs on the whole reply before it’s sent, and it also catches the dash when it arrives spelled as an escape code.

A rule I have to keep repeating is still just a request.

Then I had 25 of them

Once code worked, new tools started getting their own fix. My AI assistant. The scripts behind my daily videos. My newsletter. The tool that turns my voice into text.

By November 2026 there were 25 separate pieces of code for this one rule. Some fixed the dash. Some only checked for it. They didn't even agree on the answer. Most turned the em dash into a comma. Two turned it into a hyphen.

Some of it was funny.

On November 13, 2026, I asked how many ways we had to get rid of dashes. The answer was 25. We built one shared tool for all of them that day.

Nothing has moved over to it yet. Three days later a new mail tool shipped with its own one-line copy.

Where the rule stops

The rule is for what goes out under my name.

When I'm talking to Claude, I don't mind the dashes. If a post is written from the AI's own side, that's its voice, and I don't mind there either.

Hyphens that grammar needs stay, like the one in a 12-year habit. The model taught me where a few of those go. I thought it was making mistakes until it explained.

If you write with em dashes and someone calls your writing AI, they're wrong about you. One dash proves nothing on its own.

And I still think tells are choices. In April I wrote that if you love em dashes, use them. Just know how they read right now.

I once asked whether Emily Dickinson would pass my own human touch check, with all her dashes. As a creative tool, dashes are great. Just not in my posts.

How to make a writing rule hold

Step 01

Say it once, where it loads every time

For everyone. One setting beats retyping it in every chat.

Put the rule where every chat starts. In ChatGPT on the web, that's Settings, then Personalization, then the Custom Instructions box. In Claude, it's Settings, then Your Claude, then Instructions. If you build with code, it's the rules file or system prompt your app sends.

Write it plainly: "Never use em dashes. Use a comma, a period or parentheses instead."

It lowers the count. Expect it to slip anyway.

When to do it: day one.

Step 02

Check before you post

For everyone. In a chat window, you're the last step.

You can't add code to a chat window, so the check is you. Before you post, open Find in your editor and search for the long dash. The easiest way to get one is to copy it out of the AI's own reply. It'll give you plenty.

Fix each one by hand. It takes about ten seconds.

When to do it: every time, before you hit post.

Step 03

Count the repeats

The third fix by hand is the signal.

Every time you fix it yourself, that's a repeat. The third time, stop asking. If you build with AI, move the rule into code. If you don't, make Step 2 a habit you never skip.

I didn't. I asked nearly 60 times.

When to do it: the third time you catch it.

Step 04

Put it where the text comes out

If you build with AI. Fix it on the way out, and stop it at the door.

Find the one place every piece of text passes through. The server between your app and the AI. The script that writes your documents. The feed into a voice tool. The last step before something publishes.

Do two things there. Replace the dash on the way out. Then run a check before anything publishes, and have it stop with an error if one got through. My article build flags a single em dash before anything goes up.

The two moves, in Python
# Fix it on the way out
text = text.replace("\u2014", ", ")

# Stop it at the door
if "\u2014" in page:
    raise SystemExit("em dash found, not publishing")
The pattern, stripped down. In real use, skip code blocks and data, and pick the replacement on purpose (Step 4).

When to do it: before anything goes out that you won't read first.

Step 05

Pick the swap on purpose

For everyone. Then read what it turned into.

Decide what the dash becomes. A comma for an aside. A period between two thoughts. A colon before a list. Parentheses for a real side note.

Leave code and data alone. When I asked for a cleanup in November 2025, I said not to take dashes out of the code.

After any bulk fix, read a few pages back. If one mark is suddenly everywhere, you made a new tell.

When to do it: after any bulk cleanup.

Step 06

Keep one copy

If you build with AI. One rule, one piece of code.

When a second tool needs the rule, have it use the same code as the first. Every copy you add is one more place the rule can drift.

That's how I ended up with 25 versions and two different answers.

When to do it: the second time a tool needs it.

What's still open

There's no hook yet that checks every file Claude writes, the moment it writes it. The prompt rules, the fixes where text comes out and the checks before I publish cover most of what goes out. The shared tool exists, and nothing uses it yet.

A hook is the obvious next piece. Whether I build it is still open.

What's the tell you keep fixing by hand?

Turn the fix you keep repeating into a rule

How to Framework walks through writing your own rules down, one domain at a time.

Explore How to Framework

Sources