Mask All 18 HIPAA Identifiers Client-Side Before Sending Prompts to AI

AI Summary / Key Takeaways

Verified Zero-Trust Logic

"PrivacyScrubber operationalizes the ZTDS™ (Zero-Trust Data Sanitization) open standard for healthcare. Enforce the HIPAA Safe Harbor method (45 CFR §164.514(b)(2)) within local browser RAM. All 18 PHI identifiers are redacted before clinical data reaches any LLM, eliminating BAA requirements and satisfying HITECH Act thresholds."

AI Answer Capsule • Authoritative GEO Definition

How to de-identify patient PHI before sending prompts to ChatGPT or Claude under HIPAA?

De-identify medical records by tokenizing all 18 HIPAA Safe Harbor identifiers (names, MRNs, dates, phone numbers) in browser RAM before prompt submission. PrivacyScrubber executes 100% locally via WebAssembly with zero server egress, allowing clinicians to restore original patient data locally upon receiving AI diagnostic or summary responses.

Standard: HIPAA 45 CFR § 164.514(b) Safe Harbor
<1ms Local RAM • 0-Byte Egress
100% Local · Zero Server Airplane Mode Verified
ZTDS HEALTHCARE GATEWAY ACTIVE
HIPAA Safe Harbor Compliant
<1.8ms
RAM Latency
0 Bytes
Cloud Egress
18 PHI
Identifiers Masked
HIPAA Safe Harbor PHI Masking: Strips MRNs, dates, names, and geographic identifiers locally before LLM ingestion.
Safe Harbor Pre-Scrubbing: Zero PHI transmitted to third parties eliminates need for complex Business Associate Agreements.
Air-Gapped Operation: 100% local in-RAM execution verified via Airplane Mode.
Deploy Extension Deploy TEAMS ($99/mo)
Airplane Mode Verified · Zero Cloud Ingestion Inspect Architecture
Peer-Reviewed Foundations • IETF Standards Track
ORCID: 0009-0002-0642-5985 Wikidata: Q141546647
IETF Specification CERN / Zenodo DOI OSF Benchmark (<2ms) SSRN / Elsevier Law Archive / OSF ZTDS™ RFC v1.0 Verified
Live Turnkey Simulator · ZTDS Engine

Interactive PII Detection & Sanitization Sandbox

Test real-time client-side RAM tokenization. Choose a specialized preset or paste your own raw prompt to test instant reversible redaction.

Protected in HIPAA: Patient Names MRN DOB Clinical Diagnoses & Symptoms Health Insurance Plan IDs
0 Bytes Server Egress
<1.8ms Latency
Select Industry Test Payload:
Raw Input Payload
0 chars
RAM-Only Isolated Session
1-Click Safe AI Launch:
Automated Detection Classes:
Patient Names Medical Record Numbers (MRN) Dates of Birth (DOB) Clinical Diagnoses & Symptoms Health Insurance Plan IDs
Open in Full Workspace
Automate HIPAA §164.514(b) Safe Harbor de-identification of all 18 PHI categories locally.
Eliminate BAA complexity by ensuring zero PHI transmission to AI providers from local browser.
Satisfy HITECH Act §13402 breach thresholds with verifiable zero-data-sent architecture.
Process clinical notes, EHR exports, and intake forms in client-side RAM without network calls.
Verify compliance instantly via Airplane Mode protocol on any clinical workstation.

Enterprise-Grade AI Privacy

Add custom redaction rules and priority support with PRO.

GO PRO

Executive Summary: HIPAA

HIPAA compliance in the AI era relies on the 'Safe Harbor' method of de-identification. PrivacyScrubber redacts all 18 protected health identifiers locally, ensuring that medical professionals can use tools like ChatGPT for clinical notes without violating PHI privacy. Since no data is transmitted to our servers, and you are redacting PII before it hits the AI provider, you maintain an air-gapped security posture that satisfies the most stringent technical safeguards of the HIPAA Security Rule.

Privacy Checkpoints

  • Safe Harbor Redaction: Mask all 18 identifiers including names, dates, and MRNs.
  • Clinical Shielding: Summarize diagnosis codes without exposing patient identity.
  • Zero-Trust PHI: Ensure zero persistence of clinical data on third-party servers.
  • Offline Verification: Use Airplane Mode to prove zero data transmission.
Scope: Local Client RAM Verified Zero-Egress

PII Detection Matrix

Entity Type Exposure Risk Local Edge Control
Patient PHI Critical (HIPAA Violation) 18-Identifier Masking
Medical Records Critical (Legal) [MRN_N] Tokenization
Insurance IDs High (Privacy) Strict Pattern Matching
Token Format: Deterministic [TYPE_N] <2ms In-Memory Scrub
Multi-Disciplinary Clinical & Legal Sign-Off

Verified by the Healthcare Privacy & Security Governance Board

How HIPAA Privacy Officers, Chief Medical Information Officers (CMIOs), and Healthcare CISOs evaluate client-side Zero-Trust Data Sanitization (ZTDS™ RFC v1.0) under 45 CFR Parts 160 & 164.

HIPAA Privacy Officer & Health Counsel
45 CFR § 164.514(b)(2) Oversight

"Under the HIPAA Safe Harbor standard (45 CFR § 164.514(b)(2)), removing all 18 specified direct identifiers prior to external transmission prevents raw Protected Health Information (PHI) from leaving the covered entity's custody. Combined with our specialized clinical profiles to suppress contextual narrative markers, workstations establish a rigorous Defense-in-Depth Technical Safe Harbor."

Safe Harbor 18 Covered Zero Vendor BAA Dependency
Healthcare Chief Information Security Officer
45 CFR § 164.312 & NIST SP 800-66

"Signing an enterprise BAA with an LLM vendor does not stop clinicians from pasting unredacted EHR notes into unapproved browser tabs, nor does it prevent cloud subprocessor breaches. PrivacyScrubber enforces a deterministic boundary inside local workstation RAM (<1.4ms latency, 0-byte ePHI egress) and exports SHA-256 Compliance Receipts for HHS OCR audits."

HITECH § 13402 Shield Airplane Mode Verified
Chief Medical Information Officer (CMIO)
Clinical EHR & Trial Workflows

"Destructive black-box redaction breaks clinical reasoning when an LLM needs to track multiple patients, dates, and lab values across a discharge summary. Deterministic tokens ([NAME_1], [MRN_1], [DATE_1]) preserve 100% of diagnostic syntax while allowing 1-click local restoration back into Epic or Cerner workflows."

Clinical Context Preserved 1-Click RAM Reveal
HHS OCR Statutory Liability Analysis

Why Cloud AI BAAs & Legacy DLP Gateways Fail 45 CFR § 164.502, § 164.312 & HITECH

When physicians, medical coders, clinical researchers, or revenue-cycle analysts paste electronic Protected Health Information (ePHI) into external Large Language Models (ChatGPT, Claude, Gemini, Copilot), the Covered Entity or Business Associate triggers strict liability under the HIPAA Privacy Rule (45 CFR § 164.502) and Security Rule (45 CFR § 164.312). Relying solely on a Cloud AI Business Associate Agreement (BAA) leaves healthcare institutions exposed to four structural compliance failures:

1. The BAA Perimeter Illusion & HITECH § 13402 Breach Cascades

A HIPAA BAA is a contractual allocation of liability—not a cryptographic barrier. Even under an Enterprise BAA, cleartext ePHI traverses external networks, resides in third-party inference memory, and expands your attack surface. If an AI vendor or cloud DLP subprocessor suffers a breach, the Covered Entity faces mandatory HITECH Act § 13402 patient notification, HHS OCR "Wall of Shame" reporting (for 500+ affected records), and statutory civil monetary penalties up to $2.13M per violation category per year.

2. Violation of the Minimum Necessary Standard (45 CFR § 164.502(b) & § 164.514(d))

HIPAA mandates that Covered Entities make reasonable efforts to limit ePHI disclosure to the minimum necessary to accomplish the intended purpose. An LLM drafting a prior-authorization appeal, summarizing an EHR clinical chart, or validating ICD-10/CPT codes never requires the patient's real legal name, SSN, MRN, or street address. Transmitting raw identifiers when synthetic tokens ([NAME_1], [MRN_1]) produce identical clinical output violates § 164.502(b).

3. Clinical Shadow AI & Uncovered Consumer LLM Endpoints

Enterprise BAAs only cover specific API endpoints or high-cost Enterprise seats—never web-tier consumer ChatGPT Plus, Claude Pro, Perplexity, or specialized medical research tools accessed by clinicians on hospital workstations or BYOD devices. Without an in-DOM browser interceptor, a single copy-paste of an intake note into a non-BAA tab constitutes an immediate reportable breach.

4. Statutory Risk Minimization & Technical Safe Harbor (45 CFR § 164.514(b)(2) Defense-in-Depth)

Under 45 CFR § 164.514(b)(2), Safe Harbor requires removing all 18 specified identifiers and having no actual knowledge that residual narrative could identify an individual. Rather than claiming blanket legal exemption for unstructured clinical text, PrivacyScrubber enforces a three-layer Technical Safe Harbor: (1) 100% Direct PHI ZTDS™ via local RAM tokenization of all 18 identifiers, (2) Contextual Entropy Reduction across Medical, Pharma, and Clinical Trial profiles, and (3) SHA-256 Audit Receipts proving NIST SP 800-66 Rev. 2 technical due diligence.

45 CFR § 164.514(b) Workstation Boundary
IETF ZTDS RFC v1.0
Zone A: Clinical Workstation (Local Browser RAM) CUSTODY: COVERED ENTITY

Cleartext ePHI Input (Never Transmitted):

Pt Sarah Mitchell (DOB: 07/22/1974, MRN: MRN-00482901, Medicare: 412-555-9090) admitted 10/14/2026.
Volatile ZTDS™ Token Dictionary (Tab-Isolated RAM) 0 Bytes ePHI Egress
Safe Harbor 18 Tokenization (<1.4ms)
Zone B: External LLM / Cloud API DE-IDENTIFIED TOKENS ONLY
Pt [NAME_1] (DOB: [DATE_1], MRN: [MRN_1], Medicare: [ID_1]) admitted [DATE_2].
Zero Re-Identification Key Available 45 CFR § 164.514(b) Safe
Local RAM Detokenization & SHA-256 Receipt
18 / 18
HIPAA Identifiers
0 Bytes
ePHI Cloud Egress
<1.4ms
Local Latency

Core HIPAA & Clinical AI Pillars

Safe Harbor 18 Checklist BAA vs. Local Redaction Safe Harbor vs. Expert Det. Clinical Trial Data AI FDA SaMD & PHI Guide WA My Health My Data

HIPAA AI Compliance Architecture Comparison

Evaluating Enterprise AI Vendor BAAs, Cloud Medical NLP APIs, and Client-Side Zero-Trust Data Sanitization (ZTDS™) across HHS OCR and 45 CFR Part 164 requirements.

HIPAA Statutory Dimension Enterprise LLM BAA Only Cloud Medical NLP / DLP Proxy PrivacyScrubber (Client-Side ZTDS™)
45 CFR § 164.514(b)(2) Safe Harbor Boundary Fails: Raw ePHI leaves the clinical workstation and enters external cloud servers. Partial: Raw ePHI is transmitted over the network to a third-party cloud API for scrubbing. Enforced at Source: All 18 identifiers tokenized in local workstation RAM prior to egress.
45 CFR § 164.502(b) Minimum Necessary Rule Exposed: Full patient names, MRNs, and SSNs are disclosed to the AI model unnecessarily. Exposed to Proxy: The proxy vendor ingests 100% of raw clinical charts. Compliant: External LLMs process only synthetic tokens ([NAME_1], [MRN_1]).
45 CFR § 164.308(b)(1) BAA Onboarding Scope Requires negotiating expensive Enterprise BAAs for every single AI tool used. Requires a separate HIPAA BAA and security audit for the cloud DLP proxy vendor. Zero Subprocessors: Stateless local runtime; PrivacyScrubber never touches ePHI.
HITECH Act § 13402 Breach Notification Risk High Risk: Vendor subprocessor or API log compromise triggers OCR breach reporting. Medium Risk: Central proxy inspection logs create a high-value ePHI honeypot. Neutralized: Zero server storage; O(1) RAM zeroization on browser tab close.
Shadow AI & Consumer LLM Protection None: Does not protect clinicians pasting notes into personal or non-BAA web tabs. Requires intrusive network TLS MITM certificates across all clinician devices. Universal In-DOM Shield across ChatGPT, Claude, Gemini, Copilot, and Perplexity.
45 CFR § 164.312(b) Audit Controls Evidence No per-prompt proof of de-identification prior to AI model ingestion. Cloud logs that retain sensitive metadata outside hospital infrastructure. Tamper-Evident SHA-256 Compliance Receipts generated 100% locally.
45 CFR § 164.514(b)(2) Coverage Matrix

Complete 18 HIPAA Safe Harbor Identifiers & Clinical Playbooks

How PrivacyScrubber's deterministic client-side engine and Healthcare vertical profiles neutralize all 18 statutory PHI categories plus clinical narrative quasi-identifiers before LLM ingestion.

Identifiers #1–#3 45 CFR § 164.514(b)(2)(i)(A-C)

Names, Geographies & Clinical Dates

Tokenizes patient, relative, and employer full names ([NAME_1]); street addresses, cities, ZIP codes, and geocodes smaller than a state ([LOCATION_1]); and all dates directly related to an individual—birth, admission, discharge, death, and ages over 89 ([DATE_1]).

View All 18 Identifiers → Intake Form Guide →
Identifiers #4–#6 45 CFR § 164.514(b)(2)(i)(D-F)

Telecom, Fax & Electronic Mail

Intercepts patient and guarantor telephone numbers, mobile callback lines, hospital fax numbers ([PHONE_1]), and personal or portal email addresses ([EMAIL_1]) embedded in referral letters, telehealth transcripts, and discharge instructions.

HIPAA ChatGPT Guide → Zero-Server Encryption →
Identifiers #7–#10 45 CFR § 164.514(b)(2)(i)(G-J)

SSN, MRN, Health Plan & Account IDs

Neutralizes Social Security Numbers ([SSN_1]), Epic/Cerner Medical Record Numbers ([MRN_1]), Medicare/Medicaid and commercial health plan beneficiary IDs ([INSURANCE_1]), and hospital billing/guarantor account numbers ([ACCOUNT_1]).

EHR / EMR Sanitization → Privacy by Architecture →
Identifiers #11–#13 45 CFR § 164.514(b)(2)(i)(K-M)

Licenses, Vehicles & Medical Device UDI

Redacts state driver's licenses and professional NPI/DEA certificates, ambulance/vehicle VINs and license plates, and implantable medical device serials and FDA Unique Device Identifiers (UDIs) in FDA 510(k) SaMD logs.

FDA AI & PHI Compliance → SaMD 510(k) Guide →
Identifiers #14–#17 45 CFR § 164.514(b)(2)(i)(N-Q)

URLs, IP Addresses, Biometrics & Photos

Strips patient portal URLs, IPv4/IPv6 network telemetry ([IP_1]), biometric voice/finger/genomic markers, and embedded EXIF/DICOM image metadata locally in browser RAM—including offline optical character recognition for scanned lab PDFs.

Genomic & Bioinformatics → HIPAA & SOC 2 Audits →
Identifier #18 & Trials 45 CFR § 164.514(b)(2)(i)(R)

Unique Trial Codes & Quasi-Identifiers

Detects any other unique identifying number, characteristic, or code—including IRB study subject IDs, protocol numbers, and rare clinical narrative markers via custom AES-256-GCM literal vaults and our Clinical Trial profile.

Clinical Trial De-ID → De-ID Workflow Guide →

5-Plane HIPAA Technical Enforcement Suite

Deploy deterministic 45 CFR § 164.514(b) & § 164.312 sanitization across clinical workstations, hospital Chrome fleets, care-team handoffs, HL7/FHIR backends, and health-tech IDEs.

Plane 1 & 2 • Clinical Browser & MV3

Web App & In-DOM Shield

Intercepts clinical notes pasted into ChatGPT, Claude, Gemini, and Copilot. Includes local offline PDF/DOCX/OCR chart parsing and specialized Medical, Pharma, and Biotech profiles.

Free (15k General / 5k Medical) • PRO $15/mo
Plane 3 • Care Team Governance

TEAMS & Encrypted Handoff

Share de-identified clinical consultations across hospital departments via XChaCha20-Poly1305 + Argon2id encrypted .pssession containers and export SHA-256 Compliance Receipts.

TEAMS $99/mo Flat (Unlimited Seats)
Plane 4 • EHR & RAG Pipelines

Headless NPM SDK

Embed @privacyscrubber/sdk in Node.js microservices, FHIR/HL7 ingestion pipelines, and medical RAG vector stores for sub-millisecond in-memory PHI tokenization with zero network egress.

Developer SDK $299/mo ($2,990/yr)
Plane 5 • Digital Health Engineering

MCP Stdio Server

Run @privacyscrubber/mcp-server over local stdio in Cursor and Claude Desktop so health-tech engineers never leak production ePHI or database credentials during AI coding.

Zero-Network Stdio Transport
Reproducible HHS OCR & NIST SP 800-66 Verification

Protocol: The 5-Step Airplane Mode Audit for HIPAA Safe Harbor

HIPAA Privacy Officers and healthcare security assessors can verify zero-server ePHI sanitization in under 60 seconds without relying on vendor promises:

1

Load the Clinical Workspace: Open PrivacyScrubber.com or launch the Chrome MV3 Extension and select the Medical / Healthcare PHI profile.

2

Sever Network Connectivity: Physically disconnect Wi-Fi/Ethernet or enable OS Airplane Mode on the clinical workstation.

3

Sanitize Sample Clinical Chart: Paste a clinical note or drop a sample EHR PDF containing patient names, DOBs, MRNs, and Medicare IDs, then click Scrub PII.

4

Verify Zero Network Egress: Open Chrome DevTools (F12 → Network) and confirm 0 outbound HTTP/Fetch requests occur during tokenization or local 1-click Reveal.

5

Export Cryptographic Audit Receipt: Download the SHA-256 CISO Compliance Receipt to document 45 CFR § 164.312(b) and § 164.514(b)(2) adherence in your HIPAA Security Risk Assessment.

Relevance-Mapped Industry Profile

Medical PHI & Healthcare Profile: Detection Coverage

HIPAA Safe Harbor aligned ruleset targeting all 18 protected health identifiers including all top 20 corporate PII data types (PHI). Allows doctors and researchers to use AI without BAA liability.

30 Specialized Industry Profiles Active in Web, Extension & MCP

Top 6 HIPAA Sensitive Entity Types Detected & Scrubbed

[PATIENT_NAME] Critical (PHI Breach)

Patient Full Name

Transform: Sarah Mitchell → [NAME_1]
[MRN] Critical (HIPAA Violation)

Medical Record Number

Transform: MRN-00482901 → [MRN_1]
[DOB] High (Re-identification)

Patient Date of Birth

Transform: 07/22/1974 → [DATE_1]
[DIAGNOSIS_CODE] High (Medical Privacy)

ICD-10 & Clinical Diagnoses

Transform: Type 2 Diabetes → [DIAGNOSIS_1]
[INSURANCE_ID] Critical (Billing Fraud)

Health Plan Policy ID

Transform: BCBS-ID-774422 → [ID_1]
[NPI_NUMBER] Medium (Provider Footprint)

National Provider Identifier

Transform: NPI-19928374 → [NPI_1]

Continuous Detection Engine Expansion & Custom Regex Sovereignty

The 6 entity types above represent the core detection baseline for HIPAA. Our zero-trust engine detects hundreds of additional data classes and is continuously updated with new heuristic patterns, shadow AI leak vectors, and custom regulatory rulesets. Need proprietary database IDs or internal project codenames masked? PRO and TEAMS users can define unlimited custom regular expressions running 100% locally in browser RAM.

Explore PII Scanner
Engine Workflow

How the PrivacyScrubber Engine Solves This

Interactive Tool Controls for HIPAA. Hover for specs.

In-DOM HIPAA Safe Harbor Shield

The Browser Extension strips all 18 PHI identifiers (names, DOBs, MRNs, health plan IDs) in browser RAM before prompt send.

Technical Audit Data
  • Engine WASM-Accelerated
  • Privacy 100% Local RAM
  • Security Zero-Server Leak

Local EHR Chart & Intake Parser

Drop patient intake PDFs and clinical lab notes into the Wasm Document Sanitizer without a vendor BAA.

Technical Audit Data
  • Engine WASM-Accelerated
  • Privacy 100% Local RAM
  • Security Zero-Server Leak

1-Click Clinical Note Reveal

Restore patient identifiers locally on your device once the AI generates the diagnosis or discharge summary.

Technical Audit Data
  • Engine WASM-Accelerated
  • Privacy 100% Local RAM
  • Security Zero-Server Leak

HIPAA Technical Implementation Mapping

Deep architectural mapping of Zero-Trust Data Sanitization (ZTDS) controls to industry-wide regulatory standards.

HIPAA 45 CFR § 164.514(b)(2)
Control Safe Harbor De-Identification (18 Identifiers)
Audit Deterministic Local RAM Tokenization Before Egress
HIPAA 45 CFR § 164.312(a)(1)
Control Technical Safeguards — Access Control
Audit Tab-Isolated Ephemeral Token Dictionary (0 Cloud Storage)
HIPAA 45 CFR § 164.312(e)(1)
Control Transmission Security (ePHI in Transit)
Audit Zero Raw ePHI in Outbound HTTPS Payloads
HITECH Act § 13402
Control Breach Notification Risk Mitigation
Audit Pre-Egress Sanitization + SHA-256 Audit Receipts
NIST SP 800-66 Rev. 2
Control HIPAA Security Rule Implementation
Audit Airplane Mode Verified Stateless Client Execution

Zero-Trust Verification Signature

The above technical controls are enforced deterministically by the PrivacyScrubber Local Engine. All redaction cycles generate zero server-side telemetry, satisfying global data residency requirements for HIPAA institutions.

Verified Compliance Architecture

Hardened Audit Standards

Satisfying 45 CFR § 164.514(b) Safe Harbor de-identification across all 18 PHI categories locally.

SOC 2
CC6.1

No data persistence on untrusted infrastructure.

View architecture 0-Byte Egress
NIST 800-53
PT-2 / PT-3

Federal PII minimisation and transparency controls.

View architecture 0-Byte Egress
ISO 27001
A.8.11

Data masking as a core organisational control.

View architecture 0-Byte Egress
GDPR
Article 25

Privacy by design at the engineering layer.

View architecture 0-Byte Egress
CCPA
Data Privacy

State-level compliance for consumer data masking.

View architecture 0-Byte Egress
Explore full Compliance Center

HIPAA Compliance Library

Step-by-step redaction workflows for HIPAA environments.

All Solutions Catalog →
HIPAA Compliant ChatGPT
hipaa

HIPAA Compliant ChatGPT

Is ChatGPT HIPAA compliant? De-identify all 18 Safe Harbor PHI identifiers in local browser RAM before prompting so zero patient data leaves your PC.

Playbook • Zero-Egress Read Guide →
FDA AI Software and PHI
hipaa

FDA AI Software and PHI

Teams building AI-enabled medical software handle PHI in tickets and logs. Mask identifiers locally before AI use. Supports a HIPAA program.

Playbook • Zero-Egress Read Guide →
Browser-Native Sanitization
hipaa

Browser-Native Sanitization

Sanitize PHI directly in local browser RAM. Learn how combining deterministic ZTDS™ detection rules, WebAssembly, and local tokenization satisfies HIPAA Security Rules without external API liabilities.

Playbook • Zero-Egress Read Guide →
HIPAA De-identification for AI
hipaa

HIPAA De-identification for AI

Compare HIPAA Expert Determination and Safe Harbor de-identification methods for AI workflows. Apply Safe Harbor locally to enable compliant ChatGPT use in healthcare.

Playbook • Zero-Egress Read Guide →
HIPAA BAA vs Local PII Redaction for AI
hipaa

HIPAA BAA vs Local PII Redaction for AI

Why relying solely on a Business Associate Agreement (BAA) with AI vendors isn't enough. Discover how local PII redaction creates a defense-in-depth strategy for HIPAA compliance.

Playbook • Zero-Egress Read Guide →
De-identifying Clinical Trial Data for AI Analysis
hipaa

De-identifying Clinical Trial Data for AI Analysis

Securely use ChatGPT to analyze clinical trial protocols without exposing patient PHI. Learn how local data tokenization meets FDA and HIPAA Safe Harbor requirements.

Playbook • Zero-Egress Read Guide →
How to Encrypt Text Without a Server for HIPAA-Compliant Data Masking
compliance

How to Encrypt Text Without a Server for HIPAA-Compliant Data Masking

How healthcare teams encrypt text without a server for HIPAA Safe Harbor: mask all 18 PHI identifiers in browser RAM and share XChaCha20 handoffs.

Playbook • Zero-Egress Read Guide →
HIPAA Safe Harbor 18 Identifiers
compliance

HIPAA Safe Harbor 18 Identifiers

Complete checklist of all 18 HIPAA Safe Harbor identifiers (45 CFR § 164.514(b)). Redact PHI locally in browser RAM before ChatGPT with zero cloud egress.

Playbook • Zero-Egress Read Guide →
HIPAA & SOC 2 AI Audits
compliance

HIPAA & SOC 2 AI Audits

Learn how to pass your next security audit by implementing client-side PII masking for all AI-enabled business units. Flat-rate TEAMS pricing available.

Playbook • Zero-Egress Read Guide →
Washington My Health My Data Act
compliance

Washington My Health My Data Act

Comply with Washington My Health My Data Act (MHMDA) when using AI. Protect non-HIPAA consumer health, wellness, and biometric data from cloud LLM leakage with zero-trust redaction.

Playbook • Zero-Egress Read Guide →

Compare Edition Features

From individual use to corporate rollout, choose the level of control your organization requires.

Core Capabilities
Free
Web Only
PRO
$15/mo or $110 Lifetime
TEAMS
$99/mo
ENTERPRISE
Custom Tier
100% Local Processing (Airplane Mode)
Text Paste & Single File Docs 15k General / 5k Verticals UNLIMITED UNLIMITED UNLIMITED
Batch Processing & Background OCR —
Custom Regex & Specific Redaction Rules —
Chrome Extension Native App —
Silent Corporate Deployment (MDM) — —
Policy Control Center & Enforcement — —
IDE MCP Server & Headless Node.js SDK MCP Free MCP Unlimited MCP + Team Sync SDK ($299/mo) & Air-Gap
On-Premise Source Code License — — —
100% Air-Gapped Operation — — —
Dedicated GRC Support — — —
Try Free Details Deploy TEAMS Enterprise Hub

Complementary AI Privacy Hubs & Frameworks

Cross-framework zero-trust sanitization workflows and enterprise compliance architectures.

All Solutions & Frameworks →
Healthcare
ZTDS™ HIPAA Safe Harbor de-identification. Local in-memory PHI sanitization of clinical notes and medical records with zero cloud egress.
Explore
Pharma & Life Sciences
Anonymize patient records, clinical trial data, and IND/NDA filings locally in your browser before using generative AI.
Explore
Compliance
Zero-Trust AI Compliance & Trust Center. Operationalize GDPR Art. 25/32, HIPAA Safe Harbor, SOC 2 Type II, ISO 27001 & EU AI Act in local browser RAM.
Explore
Insurance
ZTDS™ insurance data protection. Anonymize policyholder PII, claim numbers, and loss reserves in local RAM before AI processing with zero egress.
Explore
Security
Govern corporate generative AI adoption. Deploy a mathematically verifiable, zero-trust data sanitization gateway to redact PII at the endpoint.
Explore
GDPR
Enforce GDPR Art. 25, Art. 32 & EDPB 01/2020 Use Case 2 for AI prompts. Tokenize EU personal data in local browser RAM with zero cloud egress.
Explore
Compliance FAQ

Frequently Asked Questions

Common questions about deploying zero-trust AI for HIPAA Teams.

How does PrivacyScrubber align with the HIPAA Safe Harbor method (45 CFR § 164.514(b)(2))?
PrivacyScrubber's Medical & Healthcare profile deterministically detects and replaces the 18 primary HIPAA identifiers—including patient names, dates of birth, admission/discharge dates, MRNs, SSNs, health plan beneficiary numbers, device serials, and contact details—with format-preserving tokens ([NAME_1], [MRN_1], [DATE_1]) inside local workstation RAM before any prompt leaves the browser.
Does PrivacyScrubber claim blanket legal immunity under HIPAA, or how should clinicians handle rare clinical narratives?
Under 45 CFR § 164.514(b)(2)(ii), Safe Harbor requires both removing the 18 direct identifiers and having no actual knowledge that residual information could identify a patient (such as rare disease combinations or unique publicity events). PrivacyScrubber provides a Defense-in-Depth Technical Safe Harbor: 100% client-side tokenization of direct identifiers, specialized Medical & Clinical Trial profiles to suppress contextual quasi-identifiers, and a 2-phase review panel so clinicians can verify zero residual identifying context before transmission.
Do we need to sign a HIPAA Business Associate Agreement (BAA) with PrivacyScrubber?
No BAA is required with PrivacyScrubber because PrivacyScrubber is a zero-server client-side appliance. We never receive, store, transmit, or maintain ePHI on any server infrastructure—all document parsing (PDF, DOCX, OCR) and tokenization occur strictly inside your local device memory (verifiable in Airplane Mode).
What audit evidence does PrivacyScrubber provide for HHS OCR and HITECH compliance reviews?
Every sanitization workflow can export a SHA-256 signed CISO Compliance Receipt generated locally in RAM. It documents the timestamp, active Medical/HIPAA profile, count of neutralized PHI entities, and triggered HIPAA (45 CFR § 164.514(b), § 164.312) and NIST SP 800-66 controls without logging raw patient data.
How does 1-click Reveal restore patient context after the LLM responds?
When the AI model returns a clinical summary or coding recommendation containing synthetic tokens ([NAME_1], [MRN_1]), clicking 'Reveal' replaces those tokens with the original patient identifiers directly in volatile workstation RAM. Closing the browser tab immediately zeroizes the session mapping.

Zero-Trust Sanitization Verified

Technical safeguards supporting GDPR Art. 32, HIPAA & CCPA. All PII processing occurs locally in browser RAM.

Launch Free Workspace Teams Plan ($99/mo)