"It's not just an AI writing tell - it's practically a confession."
That sentence you just read uses the same construction that large language models produce constantly. The phrase "it's not just X, it's Y" has become so saturated in AI-generated content that editors now treat it as near-definitive proof of machine authorship - on par with the notorious overuse of "delve," "tapestry," or "nuanced."
The pattern has appeared in professional financial journalism, including Barron's, making it a documented diagnostic marker for AI text - not merely a suspicion, but something approaching a guarantee.
The Rhetorical Move LLMs Can't Stop Making
This isn't random. The "not just X, it's Y" structure is a rhetorical escalation device. It takes a modest point and inflates it into something that sounds more profound - the second half "corrects" the first, implying the author is upgrading an incomplete understanding.
Human persuasive writing uses this occasionally and deliberately. Language models use it reflexively, across every subject, at volume. Consider how often you've seen sentences like:
- "It's not just a tool for writing - it's a new way of thinking."
- "It's not just about saving time - it's about reclaiming your creative energy."
- "It's not just a chatbot - it's a strategic partner."
These are structurally identical. They all inflate a real observation into something vaguer and more grandiose. And they all read the same way to anyone who has spent time editing AI drafts.
Other well-documented tells: leading with "Certainly!" or "Of course!", stacking hedge phrases ("it's important to note," "it's worth considering"), and the near-universal "delve into." Tools like ChatGPT and Claude have reduced some of these in recent model generations, but "not just X, it's Y" has proven stickier - possibly because it mimics patterns common in sales and marketing copy, which is heavily represented in their training data.
Before You Hit Publish
For anyone managing AI-assisted content, this is now a practical QA step, not an abstract concern. Search your drafts for "not just" before publishing. Most instances are pure rhetorical padding - the sentence works better with the second half alone.
"It's not just a time-saver - it's a strategic asset" says nothing that "it's a strategic asset" doesn't say more directly. Cutting the first half makes the sentence stronger and the writing more honest.
The harder problem is systemic. New models train on content that was itself AI-generated, meaning patterns from earlier model generations get reinforced rather than diluted. The construction exists in enough AI-generated training data now that it's likely to persist across multiple future model generations, even as developers try to suppress it.
Barron's is a professional financial publication with a full editorial staff. This pattern still made it through their process. That's a useful benchmark for how normalized these verbal tics have become - and how much active vigilance it takes to catch them. A text search for "not just" costs nothing and takes five seconds.