A privacy-focused browser utility to inspect and sanitise hidden digital traces and predictable linguistic tells in AI text. What it purges: • Copy-Paste Data: Strips hidden HTML class names, zero-width characters, and exotic spacing from chat interfaces. • Stylistic Tells: Detects and normalises overused AI punctuation patterns (like em-dashes for commas). Runs 100% locally in your browser sandbox. No data leaves your machine. (Does not decode proprietary server-side watermarks).

Hey Product Hunt! 👋 I built AI Trace Purifier as an experiment following the watermarking news that came out in early August. The internet is currently flooded with tools claiming 100% cryptographic watermark removal, which is statistically impossible from the front end. This tool fixes what is actually real. When you copy out of chat interfaces, you carry hidden span classes and zero-width characters with you. More importantly, LLMs have highly predictable stylistic ticks, like overusing em-dashes where a comma belongs, or abusing specific punctuation strings. I wanted to build a verifiable, utility that strips the hidden code artifacts and highlights the stylistic tells, running 100% locally in your browser so your data stays yours. Check it out and let me know where it can be tightened!

Chrome Web Store

Sanitise AI copy-paste artifacts & stylistic text tells.

Sanitise AI copy-paste artifacts & stylistic text tells.