AITACS CRM โ€” AI Therapy Assistant Calendar Suite  |  skillbuilder.club
AITACS CRM

Your Clients' Data Is Safe

AITACS CRM is built from the ground up with privacy and compliance at its core. Here's exactly how we protect sensitive clinical information.

๐Ÿ›ก HIPAA Safe Harbor ๐Ÿ‡ช๐Ÿ‡บ GDPR Compliant ๐Ÿ” TLS 1.2+ Encryption โœ“ De-identified AI
๐Ÿšซ

Names Never Leave

Your clients' real names, contacts, and dates of birth are never sent to AI. Period.

๐Ÿ”’

Dual-Layer Protection

Two independent filters remove identifying data โ€” in your browser and on our server.

๐Ÿ“‹

Full Audit Trail

Every AI request is timestamped and marked as de-identified for compliance verification.

How De-identification Works

When you use the AI assistant, your client's data passes through an automatic de-identification process before anything reaches the AI provider. This is not optional โ€” it is enforced by the system architecture.

Data Flow: From Your Browser to AI and Back
๐Ÿ“

Client Record

Full data including name, contacts, DOB, clinical notes

๐Ÿงน

Browser Filter

Removes all 18 HIPAA identifiers. Converts DOB โ†’ age.

๐Ÿ›ก

Server Filter

Second independent check. Scrubs residual patterns.

๐Ÿค–

AI Response

Receives only pseudonym + age + clean clinical context

This architecture follows the HIPAA Safe Harbor standard (45 CFR ยง164.514(b)), which defines exactly which 18 types of identifiers must be removed for data to be considered de-identified. Once de-identified, the data is no longer classified as Protected Health Information (PHI) under HIPAA.

What the AI Sees vs. What It Never Sees

โœ— AI Never Receives

๐Ÿšซ Real names (first, last)
๐Ÿšซ Phone numbers
๐Ÿšซ Email addresses
๐Ÿšซ Physical addresses
๐Ÿšซ Dates of birth
๐Ÿšซ Emergency contacts
๐Ÿšซ Social security / ID numbers
๐Ÿšซ Photos or biometric data

โœ“ AI Receives (De-identified)

โœ“ Pseudonym (e.g., "Client-A7")
โœ“ Age (integer, not DOB)
โœ“ Gender
โœ“ Presenting complaints
โœ“ Session notes (scrubbed)
โœ“ Therapeutic goals
โœ“ Progress summaries
โœ“ Assessment results

Security Layers

๐Ÿ”

Encryption in Transit

All communications between your browser, our server, and the AI provider are encrypted using TLS 1.2 or higher. No data travels in plaintext.

๐Ÿงน

Dual-Layer De-identification

A client-side JavaScript filter runs in your browser before data leaves your device. An independent server-side PHP filter provides a second check before forwarding to the AI. Even if one layer fails, the other catches it.

๐Ÿ”‘

Authentication & Access Control

Firebase Authentication with unique User IDs. Your data is isolated โ€” no other user can access it. Two-factor authentication (2FA) is supported and recommended.

๐Ÿ“Š

Audit Trail

Every de-identified data object sent to AI is marked with a timestamp and compliance method identifier. This creates a verifiable record for regulatory audits.

๐Ÿ’พ

Data Backup & Recovery

Regular encrypted database backups. Data export available in JSON format at any time. Full deletion upon account termination within 30 days.

๐Ÿšจ

Incident Response

Documented procedures for detecting, containing, and reporting security incidents. GDPR-compliant 72-hour notification and HIPAA-compliant 60-day notification. See Incident Response Policy.

Compliance Standards

๐Ÿ‡บ๐Ÿ‡ธ

HIPAA (USA)

Data transmitted to AI is de-identified per the Safe Harbor standard (45 CFR ยง164.514(b)). Platform architecture follows HIPAA Privacy Rule and Security Rule requirements. Business Associate Agreement available for Covered Entities. Read BAA โ†’

๐Ÿ‡ช๐Ÿ‡บ

GDPR (EU/EEA)

Full compliance with GDPR Articles 5โ€“49. Data Processing Agreement (Art. 28) available. Data subject rights (Art. 15โ€“22) supported. International transfers protected by Standard Contractual Clauses. Read DPA โ†’

๐Ÿ‡ฎ๐Ÿ‡ฑ

Israeli PIPA

Compliant with Israel's Protection of Privacy Act (1981) and the Privacy Protection Regulations (Data Security). Provider is registered in the State of Israel.

About the AI Provider

AITACS CRM uses OpenAI's API for AI-assisted analysis. Key facts about how OpenAI handles your de-identified data:

๐Ÿค–

OpenAI API Data Handling

No model training: API data is not used to train or improve OpenAI's models. Retention: API inputs and outputs may be retained for up to 30 days for safety monitoring, then deleted. Remember: OpenAI receives only de-identified data โ€” pseudonym, age, and scrubbed clinical context. No PHI.

Shared Responsibility

Security is a partnership between AITACS CRM and you as the practitioner. We handle server security, de-identification, encryption, and incident response. You are responsible for device security, using pseudonyms, obtaining client consent, and following the User Security Requirements.

Have security questions?

We're transparent about everything. Ask us anything.

Contact Us
Full Compliance Documentation

Terms of Service ยท Privacy Policy ยท Data Processing Agreement ยท Business Associate Agreement ยท Subprocessor List ยท Security Requirements ยท Informed Consent ยท Incident Response ยท Copyright Policy