PIIFirewall

Are you accidentally handing your personal info to AI? PIIFirewall automatically detects and masks the personal information you type into AI chats and email (names, email addresses, phone numbers, addresses, and more) — inside your browser, before it's ever sent. ■ Privacy by Design Detection, masking, and restoration all happen entirely on your device. Your text, your attachments, and any detected personal information are never sent to our servers or any third party. You are the only one who holds your data. ■ Supported services Works with major generative-AI chat services (ChatGPT, Claude, Gemini) and email (Gmail, Outlook). ■ What you can do ・Auto-mask customer names and contacts before you paste them into a chat ・Detect and defend against malicious instructions (prompt injection) hidden in incoming email and attachments ・Restore masked values on-device whenever you need the originals ■ Known limitation (up front) Detection of English personal names is best-effort and only when your browser provides an on-device AI model; otherwise it falls back to deterministic baseline detection. Japanese names, email addresses, phone numbers, etc. are covered by baseline detection.