AI writing pattern checker

This is not an AI detector, and that's on purpose

AI detectors try to guess who wrote something. They get it wrong often enough that you can't build decisions on them. This tool does something more useful: it points at the exact patterns that make writing feel machine-made, so you can fix them. Whether a person or a model wrote the first draft stops mattering once the text reads well.

The patterns it flags

• Em dashes. Language models use them far more than most people do. One or two in a long piece is fine. Ten is a signature.
• Stock phrases. "Delve into", "it's worth noting", "in today's fast-paced world", "plays a crucial role" and dozens more. They add words without adding meaning.
• Uniform sentence length. People write a short line. Then a longer one that wanders a bit before it lands. Text where every sentence is about 18 words long reads flat.
• Passive voice. "The results were measured" hides who did the work. "We measured the results" is shorter and more believable.

Does Google penalise AI content?

Google has said it rewards helpful content however it was produced, and that using automation mainly to manipulate rankings breaks its spam policies. So the real risk isn't the tool you used. It's publishing generic text with no experience or insight behind it. Cleaning up these patterns is the first step. Adding what only you know is the second.