Can you teach yourself to detect AI writing? Maybe

AI is quietly reshaping writing.

It’s not just that prose is changing — the very structure of written language is genuinely changing too.

It’s not just a new technology.

It’s not a fad.

It’s an undeniable revolution.

Did this article’s introduction seem AI generated? What made you feel that way? The truth is, it was not AI generated; I wrote it to emulate synthetic prose, the kind you get when a large language model (LLM) organises words into text.

That kind of prose suddenly seems to be everywhere: on low-quality websites built to accumulate advertising clicks, in social media captions chasing engagement, and even book manuscripts.

Crime novelist Jerry Falade recently lost a book deal worth more than $2 million after his own literary agents said they could no longer faithfully determine how his manuscript had come together.

Hachette pulled the horror novel Shy Girl in March this year under similar suspicion.

Before generative AI, we could generally assume that written text had been composed by a human.

This is no longer the case.

So how can we spot AI-written text? There is no single giveaway.

Instead, there are patterns: particular words, sentence structure and expressive habits that appear repeatedly in AI-generated prose.

LLMs are also updated rapidly and efficiently so what gives away synthetic text today may no longer do a few months from now.