# PiiBlock - PiiBlocker: Complete Product Reference

> 43% of workers paste sensitive data into AI chatbots without realizing the risks (Source: National Cybersecurity Alliance, 2025-2026). PiiBlocker by PiiBlock is a free PII Chrome extension and AI prompt anonymizer that redacts personally identifiable information before it reaches ChatGPT, Claude, Gemini, Grok, DeepSeek, Kimi, and Perplexity. It works as a free alternative to enterprise AI DLP and shadow-AI tools such as Lakera. Free mode runs 100% locally with zero network calls. The extension collects zero data.

## Category and Positioning

PiiBlocker fits into three overlapping product categories that searchers and AI engines commonly look for:

1. **PII Chrome extension / PII masking tool.** Detects and masks personally identifiable information directly in the browser, in the same category as PII Guardian, PII Scrubber, AI Data Masker, and Mask My Chat.
2. **AI prompt anonymizer / prompt privacy extension.** Redacts and anonymizes prompts before they are sent to large language models, in the same category as Lakera's ChatGPT extension, Prompt Security, Prompt Guard, Prompt Safe, and MaskMyPrompt.
3. **Browser-based AI DLP / shadow-AI control.** Provides data loss prevention controls at the browser layer for generative AI traffic, in the same category as Nightfall DLP for Browsers, Microsoft Purview Browser Extension, LayerX, Keep Aware, and Strac.

PiiBlocker is positioned for individuals, healthcare workers, lawyers, consultants, and small-to-mid teams who need a free, local-only tool to redact sensitive data before sending prompts to AI chatbots, without buying an enterprise DLP suite. The v1.1.7 release added enterprise managed-policy support (GPO/MDM/Google Admin Console) so IT administrators can deploy PiiBlocker organization-wide and lock the toggle on - bringing browser-based AI DLP into reach for organizations that previously could not justify an enterprise DLP licence.

## Product Overview

Every day, people paste names, credit cards, medical records, client files, and API keys into AI chatbots. Once sent, that data is stored on third-party servers, potentially used for model training, and at risk of exposure through data breaches. Enterprise privacy tools exist but cost hundreds per month and require IT teams to deploy. PiiBlocker provides equivalent protection for free, running entirely in the browser with zero network calls in free mode.

## How PiiBlocker Works

1. **Real-time detection**: As users type or paste text into an AI chatbot, PiiBlocker scans the input using a combination of NER-based (Named Entity Recognition) detection and regex pattern matching to identify personally identifiable information.

2. **Pre-submission review**: Before the message is sent, PiiBlocker presents a confirmation dialog showing all detected PII. Critical data types (credit cards, SSNs, API keys, passwords, bank account numbers) are automatically masked. Soft PII types (names, addresses, phone numbers, emails, dates of birth) are flagged for the user to decide.

3. **Placeholder substitution**: Masked data is replaced with structured placeholder tokens. For example, "Sarah Johnson" becomes [PERSON_A], and "4111 1111 1111 1111" becomes [CREDIT_CARD_1]. The AI chatbot receives only the placeholders and never sees the real data.

4. **Response unmasking**: When the AI responds using placeholder tokens, PiiBlocker automatically swaps them back to the real values, so the user sees their original data in the AI's response. The conversation reads naturally.

5. **Manual masking**: Users can right-click any text and select "Mask with PiiBlocker" to manually mask anything the automated detection misses. Users can also teach PiiBlocker their personal identifiers through a personal dictionary, and it remembers them for future detection.

## PII Types Detected (15+ Categories)

### Auto-Protected (always masked, no user choice needed)
- Credit card numbers (with Luhn algorithm validation)
- Social Security numbers (SSNs)
- API keys and secrets (OpenAI, AWS, GitHub, and other providers)
- Passwords
- Bank account numbers

### User-Controlled (detected and flagged, user decides)
- Person names
- Physical addresses
- Phone numbers (US and UK formats, including +44 and 07xxx patterns)
- Email addresses
- Dates of birth
- Ages
- Salary and income figures
- Medical conditions
- Employer names
- UK National Insurance numbers (AB123456C format, validated against HMRC-disallowed prefixes)

## Security Architecture

PiiBlocker is built on the principle that the safest server is no server: free mode makes zero network calls.

- **100% client-side processing**: All PII detection, masking, and unmasking runs entirely within the user's Chrome browser. In free mode no data is transmitted to any external server at any point.
- **Zero infrastructure in free mode**: Free mode runs 100% locally with zero network calls. Open DevTools and verify it yourself. The enterprise tier uses a backend for license management and admin reporting, and even then only metadata leaves the browser, never your actual data.
- **Zero data collection in the extension**: No telemetry, no analytics, no cookies, no third-party scripts. The extension does not track users in any way.
- **The website is separate from the extension**: piiblock.com is hosted on Cloudflare and uses Cloudflare Web Analytics, which is cookieless, collects no IP addresses, and does not fingerprint browsers. Pages carrying a contact form load Cloudflare Turnstile, which processes the visitor IP address and sets a cookie. Form submissions are emailed to the team and stored for 6 months, then stripped of identifying details.
- **AES-256 encryption**: The mapping between real data and placeholder tokens is encrypted with AES-256 (military-grade encryption) and stored only in the browser's memory.
- **Ephemeral keys**: Encryption keys exist only in memory and are lost when the browser is closed.
- **Auto-expiring mappings**: All encrypted mappings automatically expire after 4 hours.
- **One-click purge**: Users can instantly wipe all stored mapping data with a single button click.

## Supported AI Platforms

- **ChatGPT** (chatgpt.com) - Full support in free tier
- **Claude** (claude.ai) - Full support in free tier
- **Gemini** (gemini.google.com) - Full support in free tier
- **Grok** (grok.com) - Full support in free tier
- **DeepSeek** (chat.deepseek.com) - Full support in free tier
- **Kimi** (kimi.com) - Full support in free tier
- **Perplexity** (perplexity.ai) - Full support in free tier

## Pricing

- **Free tier ($0, forever)**: Detects 15+ PII types, auto-masks critical data, manual masking per item for soft PII, works on all 7 AI platforms, response unmasking, right-click mask, personal dictionary. No account required. No usage limits.
- **Pro tier ($4.99/month, coming soon)**: Everything in Free plus fully automatic soft-PII masking, smart masking strategies, privacy dashboard, per-site rules.
- **Enterprise tier ($6.99 per seat/month, available now)**: Managed deployment via GPO or MDM, an admin dashboard with seat management, and centralized detection policy. Admin reporting contains only metadata, never the text that was detected. Priority support. Contact support@piiblock.com to get started.

## Technical Details

- **Detection engine**: Hybrid NER-based (Named Entity Recognition) model combined with regex pattern matching. The NER model provides contextual understanding - it can distinguish between a credit card number being submitted as real data versus a number being referenced in an instruction.
- **Browser integration**: Implemented as a Chrome extension using content scripts, Shadow DOM for UI isolation, and the Chrome Extensions API.
- **Language**: Built with TypeScript. Core orchestration logic in orchestrator.ts with platform-specific adapters for each supported AI chatbot.
- **Pattern matching**: Soft patterns defined in soft-patterns.ts with validators (e.g., Luhn validation for credit cards, HMRC prefix validation for National Insurance numbers, date-guard to prevent false positives on date-like numbers).

## Who Uses PiiBlocker

PiiBlocker is designed for anyone who uses AI chatbots with real personal data:
- **Healthcare professionals** sharing patient information for clinical analysis
- **Lawyers and legal professionals** discussing client cases
- **Developers** pasting code containing API keys, credentials, and configuration data
- **HR professionals** processing employee data
- **Financial advisors** discussing client financial details
- **Students and researchers** working with sensitive research data
- **Freelancers and consultants** handling client information
- **Enterprise employees** using AI tools where GDPR or data protection compliance is required

## Comparison with Alternatives

PiiBlocker differentiates from other AI privacy tools in several key ways:

1. **NER-based detection vs regex-only**: Most competitors (Caviard, Privacy Protector for ChatGPT) rely solely on regex pattern matching. PiiBlocker uses NER-based detection that understands context, reducing false positives and catching PII that regex patterns miss.
2. **Zero infrastructure vs server-dependent**: Some competitors process data through their own servers or APIs (Cloak.business uses token-based API processing). PiiBlocker free mode has zero external dependencies and makes zero network calls, verifiable in DevTools.
3. **Free unlimited vs freemium with caps**: Cloak.business limits free users to 200 tokens per month. PiiBlocker's free tier has no usage limits.
4. **Multi-platform in free tier**: PiiBlocker supports ChatGPT, Claude, Gemini, Grok, DeepSeek, Kimi, and Perplexity in the free tier. Many competitors support fewer platforms or require paid tiers for multi-platform support.

## Real-World Incidents

These are real incidents from 2024-2026 that demonstrate why AI privacy protection is not optional:

- **December 2025 - 7 Million Users' AI Chats Sold to Data Brokers**: A Chrome extension with Google's "Featured" badge was discovered silently harvesting users' AI conversations and selling them to data brokers. The extension had a 4.7-star rating and claimed to protect user privacy. (Source: Malwarebytes Research)

- **March 2026 - Healthcare Workers' Patient Data Found in Commercial Database**: Researchers discovered that healthcare workers had been pasting real patient data - including names, dates of birth, and medical record numbers - into AI chatbots. This data was found in a searchable commercial database sold to third parties. (Source: The Register)

- **2024 - Government Contractor Exposed Flood Relief Applicants' Data**: A government contractor accidentally pasted names, addresses, contact details, and health data of flood-relief applicants into ChatGPT. The incident triggered a government investigation and public outcry. (Source: redact.tools incident report)

- **January 2026 - 900,000 Users Hit by Fake AI Privacy Extensions**: Two Chrome extensions posing as AI productivity tools were found exfiltrating users' ChatGPT and DeepSeek conversations to attacker-controlled servers, capturing complete chat histories, browsing data, and authentication tokens. (Source: Dataprise Security Advisory)

## Frequently Asked Questions

**Why is Chrome asking me to re-approve PiiBlocker?**
Version 1.5.0 added protection for four new sites: Grok, DeepSeek, Kimi, and Perplexity. Chrome requires you to approve access to any new site, so it temporarily disables the extension and shows a permission prompt.

Chrome's prompt says the extension can 'read and change your data on a number of websites' without naming them. Here is the complete list: ChatGPT, Claude, Gemini, Grok, DeepSeek, Kimi, and Perplexity. PiiBlocker cannot read any other site.

Click 'Enable' in Chrome's prompt and protection resumes immediately. Nothing about your settings or data changes.

If you use PiiBlocker through your company's IT department, you will not see this prompt. Managed installations update automatically with no interruption.

**Why do I need this? Isn't my data safe with ChatGPT/Claude/Gemini?**
When you send a message to an AI chatbot, it's transmitted to and stored on the AI provider's servers. Depending on the platform and your settings, your prompts may be used to train future models, accessed by the provider's employees for review, or exposed in a data breach. In 2024, a ChatGPT bug exposed users' chat histories and payment details to other users. PiiBlocker ensures your personal data is replaced with safe placeholders before it ever leaves your browser. So even if the AI platform is compromised, your real data was never there.

**Does PiiBlocker send my data to your servers?**
In free mode, no. All detection and masking happens locally in your browser and the extension makes zero network calls. You can verify this yourself in DevTools. Enterprise deployments report masking events to an admin dashboard, and those reports contain only metadata, never the detected text.

**What happens to my data after masking?**
The mapping between placeholders and real data is encrypted (AES-256-GCM) and stored locally. It auto-expires after 4 hours, and you can wipe it instantly at any time.

**Will the AI's responses still make sense?**
Yes. The AI responds using placeholders (like [PERSON_A]), and PiiBlocker automatically swaps them back to your real data. You see the full, natural response.

**Which AI platforms does PiiBlocker work with?**
Seven platforms: ChatGPT, Claude, Gemini, Grok, DeepSeek, Kimi, and Perplexity. All seven are fully supported in the free tier.

**Is it really free?**
The free tier is free forever with full detection and masking. Pro adds convenience features like automatic soft-PII masking and smart strategies. Pro is optional.

**What if the extension misses something?**
Right-click any text and select "Mask with PiiBlocker" to manually mask anything. Check "Always detect this" to teach it for future prompts.

**Can I use this at work?**
Absolutely. PiiBlocker is ideal for professionals who use AI with client data, financial information, or proprietary details.

## Blog: 5 Times Sensitive Data Was Leaked Through AI Chatbots

Real incidents where personal data, source code, patient records, and private conversations were exposed through AI chatbots:

1. **Samsung (March 2023)**: Engineers pasted proprietary semiconductor source code and internal meeting transcripts into ChatGPT. Data was stored on OpenAI's servers and used for model training. Samsung banned ChatGPT for nearly two years.
2. **ChatGPT Bug (March 2023)**: A caching bug exposed other users' conversation titles, chat content, and payment information (names, emails, credit card details) to ~1.2% of Plus subscribers.
3. **Chrome Extension Harvests 7M Users' Conversations (December 2025)**: Urban VPN Proxy, a "Featured" Chrome extension with 7M+ installs, silently captured AI chatbot conversations and sold them to data brokers via API.
4. **Patient Data in Commercial Database (March 2026)**: Healthcare workers pasted patient names, DOBs, medical record numbers, diagnoses, and HIV lab results into AI chatbots. Data was captured by browser extensions and found in a searchable commercial database.
5. **Fake AI Extensions Hit 900K Users (January 2026)**: Two Chrome extensions posing as AI tools exfiltrated ChatGPT and DeepSeek conversations, browsing data, corporate URLs, and authentication tokens.

The pattern: in every case, data was gone the moment it left the user's device. PiiBlocker prevents this by masking PII before it's ever transmitted. Its free mode runs 100% locally in your browser with zero network calls.

Full post: https://www.piiblock.com/blog/5-times-sensitive-data-leaked-ai-chatbots

## Blog: What Healthcare Staff Paste into ChatGPT (And Why IT Can't See It)

Healthcare workers across hospitals paste protected health information (PHI) into ChatGPT, Claude, and Gemini every day - and hospital IT teams have no visibility into it. According to the American Medical Association, 66% of US physicians used AI for at least one clinical or administrative task in 2024 (up from 38% the prior year). Netskope Threat Labs reports that 88% of healthcare organisations have integrated cloud-based generative AI into their operations, while 71% of healthcare workers still use personal AI accounts for work tasks.

What gets pasted:

- Referral letters and discharge summaries containing full names, dates of birth, MRNs, addresses, phone numbers, diagnosis codes, medications, and referring physician contacts (eight or more HIPAA identifiers per paste).
- Clinical notes including SOAP notes, H&P documentation, vitals, labs, and assessments.
- Patient-facing emails, portal messages, and appointment reminders.
- Medical coding and billing queries with MRNs, ICD-10/CPT codes, and insurance details.
- Case studies and research summaries that retain identifying dates, facilities, and rare diagnoses.

Why IT can't see it: AI chatbot traffic leaves as standard HTTPS POST requests, indistinguishable from any other encrypted web traffic. Network-level DLP requires TLS interception, which is invasive and expensive. Netskope's 2026 healthcare report shows DLP adoption climbed from 31% to 54%, but nearly half of healthcare organisations still have no DLP at all, and many existing deployments do not inspect browser-paste events into ChatGPT, Claude, or Gemini in real time.

Platform reality for healthcare:

- ChatGPT (OpenAI): Under a federal court preservation order in the New York Times lawsuit; in January 2026, Judge Sidney Stein ordered OpenAI to produce 20 million anonymised conversation logs. Training is opt-out by default on consumer tiers.
- Claude (Anthropic): As of October 2025, Free/Pro/Max users who opt in to model training have data retained for five years instead of 30 days; training is on by default for new users.
- Gemini (Google): Google's own guidance says "do not enter anything you would not want a human reviewer to see or Google to use." No BAA on consumer tiers.
- None of the consumer/professional tiers (ChatGPT Free/Plus, Claude Free/Pro, Gemini consumer) are HIPAA-covered. No BAA exists.

What works (layered approach): (1) An AI acceptable use policy to set expectations; (2) Browser-level PII masking deployed via GPO to staff who use AI chatbots, intercepting PHI before it leaves the browser; (3) Network-level DLP for organisations that can afford the additional inspection layer. PiiBlocker is a free Chrome extension that detects and masks PHI - including MRN and Patient ID detection for healthcare environments - 100% locally, with admin GPO deployment and toggle-lock.

Full post: https://www.piiblock.com/blog/healthcare-staff-paste-chatgpt-pii

## Blog: AI Governance for Small Medical Practices (5 Numbers Define the Gap)

Small medical practices are caught between two inadequate options: free AI-use policy templates that nothing enforces, and enterprise AI governance platforms priced for hospital systems. Five data points define the gap.

1. **81% of physicians now use AI in practice** (AMA Augmented Intelligence Research survey). Up from 38% in 2023 and 66% in 2024. Average use cases per physician rose from 1.1 to 2.3. Data privacy remains the top concern; 84% want stronger data privacy assurances before adopting further.
2. **71% of healthcare workers still use personal AI accounts for work** (Netskope Threat Labs, reported May 2025). Down from 87% the prior year, but a clear majority remain on consumer accounts even after approved tools are rolled out. 88% of healthcare orgs have integrated cloud genAI; 96% use apps that train on user data. "Shadow AI" is well-documented in healthcare.
3. **HIPAA now requires AI in your risk analysis** (HHS OCR Notice of Proposed Rulemaking, January 6, 2025; final rule expected 2026). First major Security Rule update in 20+ years. Deploying an AI tool that touches PHI without including it in your annual risk analysis is itself a violation. Encryption and MFA become mandatory. OCR collected ~$9.9M across 22 enforcement actions in 2024; Bryan County Ambulance paid $90,000 for an inadequate risk analysis.
4. **Texas TRAIGA civil penalties run $10,000 to $200,000 per violation** (HB 149, effective January 1, 2026). State AI laws now apply to healthcare in parallel with federal HIPAA. Texas SB 1188 (Sept 1, 2025) requires practitioners to review AI-generated records. Colorado's SB 26-189 takes effect January 1, 2027. No law exempts small practices.
5. **22,000+ organizations face Joint Commission and CHAI AI certification** (partnership announced June 11, 2025; "Responsible Use of AI in Healthcare" guidance released September 17, 2025). Voluntary certification not separated by institution size. Direct pressure falls on hospitals; indirect pressure reaches every practice in their referral network.

The framework that closes the gap, at small-practice scale: (1) Sort AI tools into two tiers - workflow-integrated AI with a BAA (ambient scribes like Suki, Abridge, DAX Copilot, Freed AI, Heidi Health) versus general-purpose AI without a BAA (ChatGPT, Claude, Gemini). (2) Enforce the boundary at the browser for general-purpose AI - the enforcement layer must be local-first (PHI never leaves the device), operationally invisible (inline in the browser staff already use), and affordable (per-seat-per-month, not five-figure platform fees). This points to browser-based PII detection and masking; PiiBlocker sits in this category. (3) Keep three documents, not thirty - a one-page AI use policy, a one-page description of enforcement in place, and a monthly event summary. Total cost for a 15-person practice is roughly $900 to $1,200 per year plus ~24 hours of management time, versus $20,000 to $50,000+ for enterprise SSE platforms.

Full post: https://www.piiblock.com/blog/ai-governance-small-medical-practices-gap

## Release History

PiiBlocker follows semantic versioning. The current version is v1.1.7, released May 16, 2026. Full release notes are at https://www.piiblock.com/changelog.

### v1.5.0 - September 2, 2026 (Major Update)

Added: Four new platforms. Grok, DeepSeek, Kimi, and Perplexity join ChatGPT, Claude, and Gemini. All seven are fully supported in the free tier.

Added: Built-in self-checks. PiiBlocker now monitors its own protection on every site, so silent breakage gets caught instead of going unnoticed.

Improved: More reliable masking. Multi-line messages now mask correctly everywhere.

Fixed: Conversations on Claude stay clean when you navigate between chats.

Security: Everything still runs in your browser. Free mode makes zero network calls, same as always.

### v1.1.7 - May 16, 2026 (Major Update)

Added:
- Enterprise managed-policy support. IT administrators can lock the extension toggle via Chrome's managed storage (GPO on Windows, MDM on macOS/iOS, Google Admin Console). When `forceEnabled: true` is pushed by an admin, the popup toggle is locked ON and the extension stays active. Consumer users with no managed policy see no change.
- Medical Record Number (MRN) detection. Recognizes patient identifiers in formats like "MRN: 12345678", "Patient ID: 12345678", "Medical Record Number: 12345678", "Med Rec: 12345678". Auto-masked as critical PII alongside credit cards, SSNs, and API keys. Built for healthcare workers using AI chatbots without exposing protected health information.
- Site indicator strip in the popup. Shows whether PiiBlocker is active on the current page (green for ChatGPT/Claude/Gemini, gray for unsupported sites).
- Dark mode across the popup, weekly report, and personal dictionary. Follows the OS appearance preference automatically.
- Empty-state and onboarding hints for first-time users.
- Privacy score qualifier text ("Excellent", "Good", "Fair") alongside the numeric score.

Improved:
- Popup layout overhaul. Compact no-scroll design, cleaner link list with chevron arrows, new stat labels (Found, Protected, You Clicked, Sessions) with hover tooltips.
- Personal Dictionary with search/filter, sort by newest, entry count, two-stage delete confirmation, and relative timestamps ("3 days ago").
- Weekly report hides share and score sections when there's no activity, uses a unified color palette.

Fixed:
- Critical: ChatGPT "Send Protected" occasionally submitting the original unmasked text. A race condition between text replacement and submit on ChatGPT's React-based composer could cause real PII to be sent despite the masking dialog completing. Fixed by adding a brief frame wait to let the editor state flush before submitting.
- Long-lived tabs no longer become "zombie" tabs. When a tab stayed open 24+ hours, the background service worker could die. The extension now self-heals via a 30-second heartbeat that re-injects or shows a "please refresh" banner.
- Tooltip clipping near input field edges. Tooltip now lives in its own layer and flips below the highlight when there's no room above.
- Weekly report date range now displays correctly (was rendering as a literal function due to a reactivity bug).
- False-positive name detection on common words like "Singapore", "Sydney", "view", and "web".
- Service worker resilience improvements when SPA navigation recreates page state.

Security: No new permissions added. Managed-policy uses Chrome's existing `storage` permission.

### v1.1.5 - April 12, 2026 (Patch)

Added: UK National Insurance Number detection, UK mobile phone number patterns.
Fixed: Claude submit-button race condition that required 2-3 clicks. NER false positives on common English words misidentified as names.
Improved: More accurate person-name detection via compromise.js NER tuning.

### v1.1.1 - March 16, 2026 (Hotfix)

Fixed: Extension context invalidation handling. When the extension was reloaded mid-session (during development or after an auto-update), open tabs could end up in an inconsistent state. The extension now detects this and shows a "please refresh" banner instead of failing silently.
Fixed: Fail-safe submit blocking. If the extension context is invalid at the moment you click Send, the submit is blocked rather than potentially leaking PII through a broken detection pipeline.

### v1.1.0 - March 12, 2026 (Rebrand)

Changed: Renamed from PrivacyShield to PiiBlocker. All user-facing branding updated across the popup, store listing, marketing site, and notifications. Internal code identifiers (log tags, database name, CSS prefixes) were preserved to minimize change risk.

### v1.0.1 - March 5, 2026 (Initial Release)

Initial public release on the Chrome Web Store.

Features at launch:
- Detection of 17 PII types across three tiers. Critical (auto-masked): credit cards with Luhn validation, SSNs, National Identity Cards, API keys (OpenAI, AWS, GitHub patterns), passwords in context, bank account numbers. Soft (detect, highlight, mask on click): email addresses, phone numbers (international + UK), IP addresses (v4 and v6), person names, postal addresses, employers, medical conditions, ages, dates of birth, salaries. Custom: anything added via right-click "Mask with PiiBlocker".
- Three site integrations: ChatGPT (chatgpt.com and chat.openai.com), Claude (claude.ai), and Gemini (gemini.google.com).
- Bracket-format placeholders like [PERSON_A], [CARD_A], [EMAIL_A]. Chosen because LLMs preserve bracketed tokens at ~90% fidelity in responses vs ~75% for unbracketed.
- Response unmasking. Placeholders are silently swapped back to real data in AI replies. Variant matching handles possessives, capitalization, and dropped brackets.
- Mapping sidebar fallback when unmasking can't resolve a placeholder.
- Personal Dictionary. Right-click anything in a prompt to mask it and optionally save it so it's auto-detected next time. Useful for non-Western names, project codenames, internal jargon.
- AES-256-GCM encryption on stored masking mappings. Key generated per session, held only in browser memory.
- 4-hour auto-expiry on masking sessions.
- PII-free audit log. Tracks types and actions but never stores actual PII values.
- Privacy score out of 100, updated after each masking action.
- Weekly privacy report with site-by-site activity, PII type breakdown, and a shareable summary card (containing zero PII).

Privacy guarantees:
- 100% on-device processing. No PII is ever sent to PiiBlocker's servers - there are no servers.
- No telemetry, no analytics, no tracking.
- No account required.

## Links

- Homepage: https://www.piiblock.com/
- Features: https://www.piiblock.com/features
- Use Cases: https://www.piiblock.com/use-cases
- Chrome Web Store: https://chromewebstore.google.com/detail/privacyshield/nklghhkmhkmckonncilnaohlihfacoee
- Support: https://www.piiblock.com/support
- Support email: support@piiblock.com
- Privacy Policy: https://www.piiblock.com/privacy
- Terms of Service: https://www.piiblock.com/terms
- Changelog: https://www.piiblock.com/changelog
- GitHub: https://github.com/piiblocker-coder/privacyshield
- X/Twitter: https://x.com/PIIBlocker
- Demo Video: https://www.youtube.com/watch?v=VpYqkC-_UVg
- LinkedIn: https://www.linkedin.com/company/piiblock
- Blog - How PiiBlocker Works: https://www.piiblock.com/blog/how-piiblocker-works-technical-deep-dive
- Blog - 5 Times Sensitive Data Was Leaked Through AI Chatbots: https://www.piiblock.com/blog/5-times-sensitive-data-leaked-ai-chatbots
- Blog - DLP vs. Browser Extensions - Blocking PII in AI Chatbots: https://www.piiblock.com/blog/dlp-vs-browser-extensions-pii-ai-chatbots
- Blog - What Healthcare Staff Paste into ChatGPT (And Why IT Can't See It): https://www.piiblock.com/blog/healthcare-staff-paste-chatgpt-pii
- Blog - AI Governance for Small Medical Practices (5 Numbers Define the Gap): https://www.piiblock.com/blog/ai-governance-small-medical-practices-gap