InferShield - AI Security Guard

About

# Never Accidentally Share Secrets with AI Again

InferShield protects you from accidentally sharing sensitive information with ChatGPT, Claude, Gemini, and GitHub Copilot.

##  What InferShield Does

**Real-Time Protection:**
Before you send any message to an AI chat, InferShield scans it for sensitive data and alerts you if threats are found. You choose what to do next: Cancel, Redact, or Send Anyway.

**15+ Threat Types Detected:**
✓ API Keys (OpenAI, Anthropic, GitHub, AWS, Generic)
✓ Authentication Tokens
✓ Email Addresses
✓ Social Security Numbers
✓ Credit Card Numbers
✓ Phone Numbers
✓ IP Addresses
✓ Dates of Birth
✓ Passport Numbers
✓ Driver's Licenses
✓ Medical Record Numbers
✓ And more!

---

## 🚀 How It Works

1. **You type** a message in ChatGPT, Claude, Gemini, or GitHub Copilot
2. **You hit Send**  InferShield intercepts the message
3. **Instant scan** (< 100ms) checks for PII and sensitive data
4. **If threats detected**, a modal appears showing:
   - Risk score (0-100)
   - Threat types found (e.g., "API Key", "Email Address")
   - Three options: **Cancel | Redact & Send | Send Anyway**
5. **You choose** what to do
6. **InferShield executes** your choice

**No threats detected?** Your message sends normally, no interruption!

---

## ✨ Key Features

**🔒 Privacy-First Design**
- Your messages are scanned in real-time and immediately discarded
- Nothing is stored, logged, or retained
- Open source — audit the code yourself on GitHub

**⚡ Lightning Fast**
- Scans complete in < 100ms
- No noticeable delay
- Works seamlessly in your workflow

**🎯 Smart Detection**
- 15+ PII patterns
- Luhn validation for credit cards
- Context-aware API key detection
- Low false positive rate

**🌐 Multi-Platform Support**
- ChatGPT (chat.openai.com)
- Claude (claude.ai)
- Gemini (gemini.google.com)
- GitHub Copilot (github.com)

**🛠️ Customizable**
- Choose "Block" or "Warn" mode
- Enable/disable per platform
- Configure your own API endpoint (self-hosting supported)

---

## 🎓 Why You Need This

**Did you know?**
- ChatGPT conversations may be reviewed by human trainers
- Claude, Gemini, and Copilot store your chat history by default
- Once you send a message, you can't take it back

**Common Mistakes:**
❌ Copy-pasting code with API keys
❌ Sharing debugging logs with personal emails
❌ Asking AI to help with config files containing credentials
❌ Pasting error messages with IP addresses or tokens

**With InferShield:**
✅ You get a warning before it's too late
✅ You can redact sensitive data automatically
✅ You stay in control of what you share

---

## 🔐 Privacy & Security

**What InferShield Collects:**
- Text you're about to send (for real-time scanning only)
- API usage metadata (timestamp, risk score, threat types)

**What InferShield Does NOT Collect:**
- ❌ Your browsing history
- ❌ Messages you don't send
- ❌ Any data from non-supported websites
- ❌ Cookies or tracking data

**Data Retention:**
- Message content: 0 seconds (immediately discarded)
- Metadata: 30 days (for billing/rate limiting only)

**Compliance:**
- GDPR compliant
- CCPA compliant
- Open source (GitHub: InferShield/infershield)

**Full Privacy Policy:** https://infershield.io/privacy

---

## 🚀 Getting Started (3 Steps)

**Step 1: Install Extension**
Click "Add to Chrome" → Extension icon appears in toolbar

**Step 2: Get API Key**
1. Go to https://app.infershield.io/signup
2. Create free account
3. Generate API key

**Step 3: Configure**
1. Click InferShield icon in toolbar
2. Enter API endpoint: https://app.infershield.io
3. Paste your API key
4. Enable platforms you use
5. Save!

**Done!** You're protected.

---

## 💰 Pricing

**Free Tier:**
- 100 scans/month
- All threat detection features
- All platforms supported

**Pro ($99/month):**
- 10,000 scans/month
- Priority support
- Advanced PII patterns

**Enterprise ($499/month):**
- 100,000 scans/month
- Custom PII patterns
- SSO integration
- Audit logs

**Self-Hosting (Free):**
- Host your own backend
- Unlimited scans
- Full control

See pricing: https://app.infershield.io/pricing

---

## 🛠️ Self-Hosting Option

**Want full control?**
InferShield is 100% open source. Host your own backend:

1. Clone repo: github.com/InferShield/infershield
2. Deploy backend (Railway, AWS, Docker)
3. Point extension to your API
4. Your data never leaves your infrastructure

**Perfect for:**
- Enterprise compliance requirements
- Privacy-sensitive environments
- Government/regulated industries

---

## 🤝 Support & Community

**Need Help?**
- Documentation: https://github.com/InferShield/infershield
- Issues: https://github.com/InferShield/infershield/issues
- Email: [email protected]

**Contribute:**
- Found a bug? Open an issue on GitHub
- Want a new platform? Submit a PR
- Security disclosure: [email protected]
---

## 📊 Supported Platforms

| Platform | URL | Status |
|----------|-----|--------|
| ChatGPT | chat.openai.com | ✅ Fully Supported |
| Claude | claude.ai | ✅ Fully Supported |
| Gemini | gemini.google.com | ✅ Fully Supported |
| GitHub Copilot | github.com/copilot | ✅ Fully Supported |

**Want more platforms?** Request on GitHub!

---

## 🔄 Regular Updates

We actively maintain and improve InferShield:
- New threat patterns added monthly
- Platform support expanded
- Bug fixes and performance improvements
- Open source development on GitHub

---

## ⚖️ Legal

**License:** MIT (open source)
**Privacy Policy:** https://infershield.io/privacy
**Terms of Service:** https://infershield.io/terms
**Developer:** HoZyne Inc.

---

## 🚨 Disclaimer

InferShield helps detect common PII patterns but is not 100% foolproof. Always review what you share with AI systems. For regulated industries (healthcare, finance), consult your compliance team before using AI chat tools.

---

**Install InferShield today and chat with AI confidently.**

⭐ **Star us on GitHub:** https://github.com/InferShield/infershield

🛡️ **Privacy-first. Open source. No secrets.**