AI Vendor Transparency Package

Framework Compliance

Every standard the HSCC AI Task Group will evaluate

🏛️

NIST AI RMF
Govern · Map · Measure · Manage

✓ IMPLEMENTED

🏥

HIPAA Security Rule
Administrative · Technical · Physical

✓ EXCEEDS

⚡

HICP 2024
Critical Practices 1–10

✓ COMPLIANT

🔗

HIC-SCRiM
Supply Chain Risk

✓ LIVE MODULE

📋

HL7 FHIR R4
ONC (g)(10) Inferno 317/317

INFERNO PASS

🛡️

42 CFR Part 2
SUD Consent Enforcement

✓ ENFORCED

Where the Sector Is. Where We Are.

Source: 2025 Healthcare Cybersecurity Benchmarking Study — Censinet/AHA/HSCC/KLAS/Health-ISAC

100%

Our NIST AI RMF
(sector avg: 31%)

100%

Supply Chain Coverage
(sector avg: 52%)

317/317

Inferno Sub-Tests
FULL PASS April 2026

Zero

PHI on MyRxWallet
Servers (ZK Architecture)

HSCC 7-Phase AI Vendor Lifecycle Response

Every phase of the Health Industry Third-Party AI Risk & Supply Chain Transparency Guide (April 2026) — answered

What This Document Is

The HSCC CWG AI Task Group published this 7-phase framework in April 2026 for use by 480+ member health systems evaluating AI vendors. Every health system CIO, CISO, and procurement team will use it. MyRxWallet is the first AI-native EHR to pre-map every requirement before being asked. Use this document in your vendor due diligence review.

HSCC Requirement

MyRxWallet Answer

Use Case Justification

Can the vendor prove clinical value? Patient outcomes data? Risk-benefit analysis?

Our Answer

ONC (g)(10) Inferno-tested health record with real-time drug interaction checks (NLM RxNav), DSCSA drug provenance verification (FDA openFDA), and FDA recall alerts. Clinical value = patient safety, not productivity theater. Patient Data Royalty marketplace incentivizes data quality. Trial participants tracked via Clinical Trial NFT (FDA 21 CFR 312 + ICH GCP E6).

Vendor Due Diligence

SOC 2 Type II? HIPAA BAA availability? AI usage disclosure? Sub-processor list?

Our Answer

✓ BAA available immediately — email info@myrxwallet.io
✓ SOC 2 Type II — bundled with Schellman audit engagement (in progress)
✓ ONC Registered Developer — Inferno 317/317 PASS, April 2026
✓ UEI: MVKQW9DKKGB8 · EIN: 33-1503628
✓ Sub-processors: Zero third-party AI APIs touch PHI. Our AI (Sentinel) runs on-premises on our own VPS. NLM RxNav and FDA openFDA are government APIs (public drug data only).
✓ Zero-knowledge architecture: a MyRxWallet server breach cannot expose patient PHI.

Contract Negotiation

AI liability terms? SLAs? Data ownership clauses? Audit rights?

Our Answer

✓ Data ownership: Patient-owned NFT architecture — patients own their data, not us
✓ SLA: 99.9% uptime target, nginx + systemd watchdog on DigitalOcean VPS
✓ AI liability: Sentinel AI decisions are advisory only; clinical decisions remain with licensed providers
✓ Audit rights: Full FHIR R4 audit log at /api/v1/fhir/r4/AuditEvent
✓ FHIR export: Patient data exportable at any time via $export (HL7 Bulk Data IG v2.0.0)

Implementation

Onboarding documentation? Staff training? Integration testing?

Our Answer

✓ API documentation: myrxwallet.io/developer.html — sandbox keys issued instantly
✓ SMART on FHIR App Launch 2.0.0 — any FHIR-compatible EHR connects in minutes
✓ CCD/C-CDA importer — patients self-import from any existing EHR
✓ CMS Blue Button + Payer API OAuth integration — automated data pull
✓ TEFCA/QHIN application submitted — Health Gorilla pathway
✓ Staff training: role-based portal (Admin/Provider/Patient) with contextual tooltips

Ongoing Monitoring

Sentinel/monitoring system? Audit logs? Anomaly detection?

Our Answer

✓ MyRx-Sentinel AI Agent v2.0 — real-time HIPAA violation detection, screenshot guard, incident log
✓ Drug recall monitoring — FDA openFDA webhook integration
✓ Anomaly detection — AI daily ops agent flags abnormal lab values, drug interactions
✓ MyRx Identity Guard — breach monitoring (HIBP integration), medical ID theft detection
✓ On-chain audit trail — Hyperledger Fabric 2.5, tamper-evident, 4 channels

Incident Response

Documented IR plan? AI-specific incident types? Breach notification timeline?

Our Answer

✓ Zero-knowledge architecture: server breach = no PHI exposed (HIPAA breach threshold not met)
✓ NFT instant lock — patient can lock their health record with one tap in emergency
✓ Sentinel incident log — every HIPAA violation attempt logged with incident ID
✓ HIPAA 60-day breach notification timeline supported — built into incident workflow
✓ MyRx-DAO Governance — Sentinel AI operates under defined governance rules even during incidents

End-of-Life Transition

Data portability? FHIR export? Offboarding plan?

Our Answer

✓ FHIR Bulk Data $export — all patient data exportable as NDJSON in one API call
✓ CCD/C-CDA format export available
✓ Patient owns NFT — data access follows the patient, not the contract
✓ No data lock-in: zero proprietary formats, all USCDI v3 compliant fields
✓ GDPR Art. 17 Right to Erasure — Medical Records Sharding module supports cryptographic erasure

Zero-Knowledge Architecture — Why a Breach Is Not a HIPAA Breach

The HSCC CWG "Secure by Design and Default" standard for third-party vendors — we built it first

🏥 Legacy EHR Architecture

PHI storagePlaintext in vendor DB

Server breach resultHIPAA breach — notify patients

EncryptionAt-rest only

Data ownerVendor

Supply chain riskAny sub-processor breach = exposure

AI audit trailVendor-controlled logs

🛡️ MyRxWallet Zero-Knowledge Architecture

PHI storageAES-256-GCM encrypted, patient-keyed

Server breach resultCiphertext only — not a HIPAA breach

EncryptionField-level + HKDF-SHA256 key derivation

Data ownerPatient (NFT-bound)

Supply chain riskDrug Provenance NFT — lot-level tracking

AI audit trailHyperledger Fabric — immutable, on-chain

AI System Disclosure

Full transparency on every AI component — per NIST AI RMF Govern 1.7 requirement

AI System	Function	Training Data	Output Type	Human Override	Bias Controls
MyRx-Sentinel	Health monitoring, drug interactions, recall alerts, anomaly detection	NLM RxNav (government API), FDA openFDA (government API) — no patient PHI used for training	Advisory alerts only	✓ Always — provider reviews all alerts	Government data sources only; no proprietary training sets
Identity Verification Agent	Document + selfie confidence scoring for patient enrollment	Confidence threshold algorithm (≥75% required). No facial recognition database.	Approve/flag/reject	✓ Always — flagged cases go to admin review	Threshold-based, not ML model dependent
MyRx-Score Engine	Patient health engagement scoring (300–850 scale)	Patient's own health data only — no external benchmarking datasets	Engagement score (not clinical diagnosis)	✓ Always — advisory only, not used for clinical decisions	Score is patient-relative, not compared to demographic cohorts
Daily Ops Agent	Scheduling optimization, anomaly detection, platform health	Platform operational data only (not patient clinical data)	System alerts and admin notifications	✓ Always	No demographic variables in operational models

Ready to onboard?

We respond to enterprise inquiries within 24 hours. BAA, pilot agreement, and SOC 2 bridge letter available on request.

Request Enterprise Package Explore API Sandbox CISO One-Pager

info@myrxwallet.io · 702.546.8686 · UEI: MVKQW9DKKGB8 · EIN: 33-1503628