Lately, I've been thinking a lot about how much sensitive information we're all feeding into AI tools without really stopping to consider the consequences.

At first, it was harmless - asking ChatGPT to help draft emails, summarize articles, or brainstorm ideas. But over time, my prompts got more detailed. I started including real client names, project info, company data… even things like meeting notes or internal discussions. It felt natural, convenient, but I started thinking a lot more about my privacy

No one reads the terms of service, and even when companies say they don’t store your prompts, the truth is murky. And the more AI becomes integrated into daily workflows, the easier it becomes to overshare, especially when you're moving fast. So I built a tool to protect myself.

It’s called Redactifi. It's a browser extension that detects and redacts sensitive information from your AI prompts before they’re ever sent. It runs entirely on your device, no cloud processing, no logs, no storage. You just write your prompt, click a button, and it automatically scrubs out things like names, emails, phone numbers, company info, and more.

Curious if anyone else has been feeling the same tension - using AI more, but trusting it less?

10 free redactions per month? 10 per day seems like a much more reasonable low amount. Why is there even a limit on number of uses for local compute? Why is it impossible to see the price? I'm glad this works for you, but it seems like you tried jumping straight to the "enshititication" stage of proprietary software before luring users in.