The Founders ClubLifetime access for 25 agencies - pay once, no subscription, white-label reporting.Claim your seat
16 Signs of AI Writing (and How to Fix Them)

Most AI text isn't wrong. It's just a little too smooth, a little too tidy, and full of the same habits every model reaches for. Once you know the signs of ai writing, you can't unsee them. Below are 16 concrete tells, grouped by type, with a fast fix for each so your draft reads like a person actually wrote it.

Punctuation signs of AI writing

Punctuation is the fastest place to catch a machine. These marks travel with copied text and rarely match how a normal person types in a plain editor.

1. Em dash overuse

Models love the em dash. They drop it everywhere a comma, a period, or a pair of parentheses would do the job just as well. One or two across a whole article is normal. Two or three per paragraph is a loud tell, and it's one of the first things a reader clocks even without knowing why.

Fix: Swap most of them for commas or periods. Read more about the em dash AI tell and why it stands out so much.

2. Curly quotes and fancy punctuation

Copy pasted AI output often carries curly quotes, curly apostrophes, and the single-character ellipsis. Most people typing straight into a form or a CMS produce plain straight quotes instead. When a paragraph is full of typographer's marks that the surrounding text doesn't have, the mismatch gives it away.

Fix: Find and replace curly marks with straight ones. Kill the ellipsis character and use three periods if you really need the pause.

3. Invisible and zero-width characters

Some models slip in zero-width spaces, non-breaking spaces, and other invisible characters you can't see but a detector can read instantly. These are among the sneakiest ai tells because they survive a copy paste and sit there silently in your text, flagging it as machine made.

Fix: Run the text through a cleaner that strips hidden characters, or retype any flagged sentences by hand.

Vocabulary tells and word choice

Word choice is where AI shows its training. It reaches for the same formal vocabulary and the same safe hedges over and over.

4. Giveaway vocabulary

Certain words scream machine: delve, robust, tapestry, leverage, seamless, comprehensive, pivotal, meticulous. No single word proves anything on its own. A cluster of them in one piece is a classic pattern, because a person writing casually almost never stacks that many high-register words together.

Fix: Cut them or swap in plain alternatives. See the full list of words that give away AI writing.

5. Boilerplate transitions

"In conclusion," "furthermore," "moreover," and "it's important to note" are stock connectors models use to sound organized and formal. Real writers rarely reach for them, and almost never three in the same piece. They add length without adding meaning.

Fix: Delete the transition, or replace it with a plain one like "so," "but," or "here's why."

6. Hedging and vague qualifiers

AI hedges to stay safe. You'll see "can be," "often," "may," "in many cases," and "generally" softening almost every claim in the piece. The result reads careful and committed to nothing. A human with an opinion sounds sharper than that.

Fix: Pick a position. Replace the hedge with a direct statement or a specific number you can stand behind.

Rhythm tells: how the text sounds

Read the draft out loud. Machine text has a rhythm problem that's obvious to the ear even when every sentence is technically fine.

7. Flat sentence rhythm

Human writing has burstiness. Long sentences bump against short ones, and the pace jumps around. AI tends to produce sentences of similar length and shape, line after line after line, which reads flat and a little robotic no matter how good the words are.

Fix: Break a long sentence into two. Add a three word sentence. Let the rhythm swing.

8. The "it's not just X, it's Y" construction

This one is everywhere in chatgpt tells. "It's not just a tool, it's a partner." "This isn't about speed, it's about trust." The mirrored setup feels punchy the first time and canned by the third. Once you notice it, you'll see it in half the AI copy you read.

Fix: Rewrite it as a plain statement. Say what the thing is without the setup and the reversal.

9. Symmetrical rule-of-three phrasing

Models love triads: "clear, concise, and correct." "Plan, build, and ship." The rhythm is so balanced it feels rehearsed, and it repeats in list after list until the whole piece sounds like a slogan generator.

Fix: Use two items sometimes, or four. Break the perfect symmetry so the line sounds spoken, not staged.

Structure tells: how the text is built

Zoom out from the words and look at the shape of the page. AI builds documents in a predictable mold.

10. Uniform paragraph length and shape

AI paragraphs often run the same length, three or four sentences each, block after block down the page. Human drafts are lumpier than that. Some paragraphs are a single line. Some ramble on. The evenness itself is a signal.

Fix: Vary it on purpose. Let a paragraph be one sentence when the point is short.

11. Over-structured lists and headings

A short answer chopped into five headings, each with a bolded lead-in and a neat bullet list, is a strong tell. Real people don't format a casual reply like a corporate deck. The scaffolding is doing the thinking, and it shows.

Fix: Flatten it. Turn some lists back into prose and drop the headings the piece doesn't need.

12. Restating the prompt or the question

AI often opens by repeating the question back at you. "Great question. When it comes to knowing how to tell if something is ai written, there are several factors to consider." That echo is one of the clearest ai writing patterns, and it wastes the first sentence saying nothing.

Fix: Cut the restatement. Start with the actual answer.

13. Empty intros and conclusions

The intro warms up for a whole paragraph without landing a point. The conclusion repeats what you just read and tacks on a tidy takeaway. Both are padding that a person short on time would never bother to write.

Fix: Delete the throat clearing. Open on your strongest point and stop when you're actually done.

Other AI tells to watch for

These last signs of ai writing are the hardest to fake your way around, because they're about substance, not surface.

14. Perfectly clean grammar with no voice

Flawless grammar paired with zero personality is itself a tell. No contractions, no fragments, no asides, no jokes, no rough edges. It reads like a product manual instead of a person, and readers feel the absence even if they can't name it.

Fix: Add contractions and a little attitude. A fragment or a quick aside signals a human hand on the keyboard.

15. No specific details or lived examples

This is the biggest tell of all. AI writes about things in general because it wasn't in the room. No names, no real numbers, no small concrete moment that only someone who was there would know to include. It's confident and completely generic.

Fix: Add one specific detail per section. A real example, a date, a price, a thing that happened.

16. Decorative emoji and bold on everything

Emoji bullet points, a checkmark in front of every benefit, and bold text sprinkled across half the sentences is a common look for AI output, especially in social posts and marketing copy. The formatting is trying to sell you before the words do.

Fix: Strip most of the emoji and bold. Keep formatting for the few spots that truly earn it.

How to fix all of these at once

You can hunt these tells down by hand, and it's worth learning to. But when you need a clean draft fast, run it through our free AI humanizer tool. It strips the hidden characters, fixes the punctuation, and flags the stock words for you in seconds.

Want to go deeper? Our AI humanizer guide and our walkthrough on how to make AI writing sound human cover the edits that matter most once the obvious signs of ai writing are already gone.

Ready to remove the AI tells?
Paste your draft into the free humanizer. No signup, first 300 words free.

Frequently Asked Questions

Can AI writing be made undetectable?

You can get very close. Strip the hidden characters, fix the punctuation, swap the stock words, vary your sentence length, and add real details, and most casual readers and many detectors will read it as human.

The one thing no tool fixes for you is substance. Genuine, specific, first-hand information is what makes writing read as truly human.

Ready to remove the AI tells?
Paste your draft into the free humanizer. No signup, first 300 words free.

Do AI detectors look for these signs?

Yes. Detectors weigh many of the same patterns covered here, especially low burstiness, meaning uniform sentence length, and predictable word choices. Some also scan for invisible characters and formatting quirks that come straight from model output.

That's why fixing the surface tells and adding real variety helps on both fronts: it reads better to people and scores lower as machine text.

Ready to remove the AI tells?
Paste your draft into the free humanizer. No signup, first 300 words free.

Do em dashes mean AI wrote it?

Not by themselves. Plenty of skilled writers use em dashes on purpose. What signals AI is the frequency: two or three per paragraph, used where a comma or period would work fine.

Treat heavy em dash use as one clue, not proof. Weigh it alongside the vocabulary, rhythm, and detail tells before you decide.

Ready to remove the AI tells?
Paste your draft into the free humanizer. No signup, first 300 words free.

How can you tell if something was written by ChatGPT?

Look for the common chatgpt tells stacked together. Em dashes in most paragraphs, curly quotes, phrases like "it's not just X, it's Y," and rule-of-three lists like "clear, concise, and correct."

Add the habit of restating your question before answering, plus a tidy summary at the end, and you're almost certainly looking at model output rather than a human draft.

Ready to remove the AI tells?
Paste your draft into the free humanizer. No signup, first 300 words free.

How do I remove AI tells from my text?

Start by cleaning the punctuation and stripping hidden characters, then cut the stock words and filler transitions. Vary your sentence lengths, flatten over-built lists, and add at least one specific detail per section.

To do it fast, paste your draft into a free AI humanizer tool that handles the punctuation, hidden characters, and giveaway words in one pass, then finish the voice edits by hand.

Ready to remove the AI tells?
Paste your draft into the free humanizer. No signup, first 300 words free.

What are the signs of AI writing?

The clearest signs are punctuation habits like heavy em dash use and curly quotes, stock vocabulary like "delve" and "robust," filler transitions like "in conclusion" and "furthermore," and a flat sentence rhythm where every line runs about the same length.

Structure gives it away too: even paragraphs, over-built lists, and empty intros. The strongest tell is a total lack of specific, first-hand detail.

Ready to remove the AI tells?
Paste your draft into the free humanizer. No signup, first 300 words free.

What is the biggest AI writing tell?

The biggest tell is text that says a lot without any specifics. AI writes in generalities because it wasn't there, so you get confident prose with no names, no real numbers, and no small concrete moment a person would actually remember.

Punctuation and word choice are easier to spot, but missing detail is the one that's hardest for a model to fake.

Ready to remove the AI tells?
Paste your draft into the free humanizer. No signup, first 300 words free.

What words are AI tells?

Common giveaway words include delve, robust, tapestry, leverage, seamless, comprehensive, pivotal, meticulous, and elevate. Stock phrases count too: "in conclusion," "furthermore," "moreover," and "it's important to note."

No single word is proof. A cluster of them in one short piece is the real signal, since a person writing plainly rarely uses that many high-register words at once.

Ready to remove the AI tells?
Paste your draft into the free humanizer. No signup, first 300 words free.