Arceris Prompt Guard – AI Data Leak Protection

Detects and masks personal, health and card data before you send prompts to ChatGPT, Claude, Gemini and more. By arceris.ai

Arceris Prompt Guard checks what you type into AI tools such as ChatGPT before it is sent. If your prompt contains personal, health or payment data, it stops the send and asks you what to do: mask it, edit it, or send anyway. WHAT IT DOES - Catches sensitive data such as national ID numbers, phone numbers, emails, IBANs, credit cards, patient and diagnosis details before they are sent - Masks values without breaking their format: the first and last characters stay, the middle is randomized - Can show you the original value again in the AI's reply - Shows a live count of sensitive items while you type WHAT'S INCLUDED - 32 sites covering the most widely used AI chat tools; the full list is on the extension's Providers page - 87 built-in rules for PII, PHI and PCI data - Checksum validation for card numbers, IBANs and national IDs, so random digit strings do not trigger false alarms - Three modes: Warn, Block, Monitor only - Every rule is editable; add your own regex rules, import and export as JSON - Add the address of your own internal AI tool PRIVACY All scanning happens in your browser. Your prompts, the data found and the statistics never leave your device; the extension sends nothing to any server. No account is required. LIMITS The extension scans text you type or paste. It does not scan files or images you upload. FOR TEAMS Prompt Guard is the free browser companion to the Arceris platform. Central policy management, file and image scanning, audit logs, reporting and Shadow AI discovery are available in the Arceris platform. Learn more: https://arceris.ai/