PII protection for every AI
Stop your team leaking
into AI.
Personal data — in prompts and uploaded files — is redacted before it reaches Claude, ChatGPT, Kimi or Gemini. Your team sees the real thing. The AI only ever sees a token. And so do we — never your data.
No card · 250 free requests · nothing sensitive ever stored
Works everywhere
One shield. Every AI.
Every major model and chat tool, through one engine.
Three ways in
Switch it on in minutes.
All share the same redaction engine. None let raw PII reach the model.
One config line
Point any app or Claude Code at the shield. Covers the Anthropic API and every OpenAI-compatible model.
ANTHROPIC_BASE_URL=
piishield.aiOne command
Add the shield as an MCP server. Any agent gets redact + coverage tools — set what to mask right from chat.
claude mcp add --transport
http shield …/mcpZero setup
Our browser extension redacts in the chat box before you hit send — on claude.ai, ChatGPT, Gemini, Kimi and more.
# install & go
# nothing to configureZero-trust by design
We never see your data.
A shield nobody on the outside can see through — including us. Self-host it and we have zero access at all.
- Your prompts or messages
- Personal data — names, emails, IDs, cards
- The token↔value map (in memory for the request, then gone)
- Counts of PII masked, by type
- Timestamp, model, status — for usage & billing
- Keys encrypted at rest (AES-256-GCM)
Simple pricing
A fraction of your AI bill.
You keep your own model billing. You only pay for the shield — never more than the AI it protects.
Prefer usage-based? ~R15 per 1,000 requests.
Ready in a day
Give your team AI
without giving away your data.
Create a free workspace, get your key, and shield every AI your team touches.
Start free →