Launching today
EraseAI Firewall
Stop secrets and PII before you hit Send in Public LLMs
2 followers
Stop secrets and PII before you hit Send in Public LLMs
2 followers
A browser firewall for AI chats. EraseAI checks your prompts in ChatGPT, Claude and Gemini for API keys, passwords, card numbers and personal data before you send. Free on-device, no account; Personal adds redaction and attachment scanning.
Free Options
Launch Team
Atlassian LoomTalk it through in Loom. Ship with AI
Promoted
Hey Product Hunt 👋
We built EraseAI Firewall after watching smart people paste production keys and customer data into AI chats — not carelessly, just fast.
What it does: when you press Send in ChatGPT, Claude, Gemini or Replit, it checks the message. If it finds a key, password, card number or personal data, it shows you exactly what, and you choose to Send Anyway or Cancel ('Personal' package adds Sanitize & Send).
What we're proud of:
Free, and it works the moment it's installed — checks run on your device, no account needed
Personal ($5/mo) also reads attachments: PDFs, Office files, screenshots (OCR in the browser) and zips
IT can roll it out to a whole company through Chrome policy
It stays out of the way when there's nothing to flag
We're a small team and read every comment — tell us what it flagged that it shouldn't have, and what it missed.