If one sentence shape gives away machine-written prose in 2026, it is this one: "It's not a tool, it's a mindset." "This isn't about speed. It's about trust." Language models reach for it constantly because it makes an ordinary claim feel like a reveal.
What it looks like
It's not X, it's Y.
This isn't about X. It's about Y.
Not just X, but Y.
X isn't the goal. Y is.
Less about X and more about Y.
A clipped tail: "The options come from the item you picked, no guessing."
Why it reads as AI
The negative half names something nobody claimed. Nobody said your onboarding was "just a checklist," so denying it adds no information. It only borrows drama. One of these in a long essay is fine. Three in a LinkedIn post and readers stop trusting the writer.
Examples and fixes
BeforeOur platform isn't just a CRM. It's a growth engine.
AfterOur CRM scores leads and tells reps which ones to call first.
BeforeThis isn't about working harder. It's about working smarter.
AfterBatching email into two blocks a day saved our team about five hours a week.
BeforeIt's not a bug, it's a feature.
AfterKeep it if the joke is the point. In product copy, say what the behavior does.
How to fix it
Delete the negative half and read what's left.
If what's left is vague ("a growth engine"), replace it with the concrete thing the product does.
Keep the contrast only when the reader really believes the negative half and you are correcting them.
When it's fine
Real corrections are fine: "The outage wasn't caused by the deploy; it was a DNS expiry." Someone thought it was the deploy, and you're telling them otherwise.
Check your own text for this pattern
Paste text to count matches.
Runs in your browser; nothing is uploaded. For the full report with all patterns highlighted, use the SlopScore checker or score a website.