Anti-AI Style Guard

A catalog of AI writing tells and a scan that removes them before publishing, so nothing you ship smells like a machine.

The Problem

Readers have learned to smell AI writing. The em dashes, the "not X but Y" constructions, the inflated superlatives and the tidy little summaries at the end of every paragraph. One synthetic paragraph can undo the trust a brand spent months building, and most people can't name the patterns they're reacting to.

The Solution

A documented catalog of the patterns that make writing read as machine-made, paired with a scan that checks every draft against it before publication. The tells get caught and rewritten while the piece is still a draft.

How It Works

We maintain a catalog of AI writing patterns: the constructions, the filler words, the fake structures and the canned phrases.
Each pattern gets a wrong and right example, so the fix is mechanical instead of a judgment call.
The catalog installs as a hard rule set in every writing tool, and drafts avoid most patterns from the start.
Before anything publishes, a scan runs the full checklist and flags every remaining hit for rewrite.
New patterns join the catalog as the models evolve, because the tells change with every generation.

Built With

A pattern catalog with paired wrong and right examples
A pre-delivery scan run against every outgoing piece
Hard-rule installation across your writing tools

It sits on top of the Writing Style rulebook and protects the Personality Definition's work, the last gate before anything carries your name.

What You Get

Published writing that passes the sniff test with readers who know AI when they see it
A zero-hit standard you can actually verify, instead of a vague feeling
A catalog that keeps pace as the patterns evolve

The Numbers

1

Day to install catalog and scan

~1

Hour back every week

~1

Week to pay for itself

One day to install the catalog and the scan. It returns about an hour a week of manual de-AI-ing, and the reputational insurance is the real product.