What this checker looks for
Language models have recognisable habits. None of them proves anything on its own, and good human writers use every one of them occasionally. What stands out is density: the same habits stacked paragraph after paragraph. The checker measures five signals.
- Stock phrases and overused words. Words and phrases language models lean on heavily, such as "delve into", "a testament to", "navigate the complexities" and "in today's fast-paced world". One is harmless. A cluster makes text read as generic.
- Uniform sentence length. People write in bursts: a long sentence, then a short one. Model output tends to hold a steady rhythm. The checker measures how much your sentence lengths vary.
- Filler transitions. Sentences that open with "Furthermore", "Moreover", "Additionally" or "In conclusion". They signal structure without adding meaning.
- Contrast framing. The "it's not just X, it's Y" construction. Effective once; a tell when it repeats.
- Dash density. Em dashes are a legitimate tool, but a high rate of them per 100 words is one of the most commented-on habits of AI text.
Why this isn't an AI detector
AI detectors promise a verdict: this text is 87% AI. The evidence says they can't deliver it. OpenAI withdrew its own classifier in 2023 because it was not accurate enough. A 2023 Stanford study found that popular detectors labelled more than half of essays by non-native English writers as AI-generated. Students and freelancers have been accused on the strength of these scores.
The underlying problem is simple. AI text is trained on human text, and humans are now learning from AI text. The two overlap, and a light edit moves either one across any line a detector draws. So this tool answers a question it can actually answer: does this read as generic, and what exactly makes it read that way?
That's useful whichever side you're on. If you drafted with AI, it shows what to rewrite so the result sounds like you. If you wrote every word yourself, it shows what might get you wrongly flagged.
How to make writing sound less generic
Replace the stock phrase with the specific claim it was standing in for: "a testament to our commitment" usually means one concrete thing you did. Vary sentence length on purpose. Cut openers like "Furthermore" and let the logic carry the reader. And say the point first. Generic writing warms up; good writing starts.
Then check the result with the Readability Checker, since plainer usually also means easier to read.
