Use case

Everything you'd use ChatGPT for —
without the HIPAA violation.

Wren is a HIPAA-compliant clinical AI assistant built on Claude. Ask it anything, bring in real session detail, and let it help — safely, under a signed BAA. The AI you wanted ChatGPT to be, made for therapists.

Try Free for 30 Days

No credit card · Cancel anytime

Why you can't just use ChatGPT.

ChatGPT and Claude's consumer apps won't sign a BAA with you, so pasting client information into them is a HIPAA violation — full stop. Wren runs on the same caliber of model (Claude, via AWS Bedrock) under a signed BAA, so you get the capability without the risk.

What therapists actually use it for.

  • Talk through best practices, or get a second perspective on a session
  • Think through a tricky case or differential, with real detail
  • Draft anything — notes, letters, treatment plans, emails, handouts
  • Brainstorm interventions, homework, or how to phrase something hard
  • Ask clinical and documentation questions and get a straight answer

Why therapists choose Wren over ChatGPT

✓
Actually HIPAA-compliant — signed BAAs with you and every subprocessor; ChatGPT and Claude's consumer apps won't sign one.
✓
The same caliber of AI — built on Claude via AWS Bedrock, the model many clinicians find writes the most naturally, so you're not trading power for compliance.
✓
More than a note-writer — talk through best practices, get a second opinion on a session, draft a hard email, or work through an ethical question — with real detail, safely.
✓
Built for therapy by a therapist — Steven Wallace, LMHC, wrote both the clinical logic and the code, so it understands documentation and clinical nuance, not just chat.
✓
Your data stays yours — never used to train any model; transcripts auto-delete on your schedule.
✓
Pay for what you use — no per-seat subscription; a typical request costs pennies, every cost visible.
Learn more about Wren →

FAQ

Is ChatGPT HIPAA-compliant for therapists?
No. OpenAI and Anthropic don't sign Business Associate Agreements with individual clinicians through their consumer apps (ChatGPT, Claude.ai), so entering client information is a HIPAA violation. Wren provides the same kind of AI under a signed BAA, which makes it compliant for clinical use.
Isn't Wren just ChatGPT with a wrapper?
No. Wren runs on Claude via AWS Bedrock under a signed BAA, adds clinical guides so it writes in your style, and includes document tools, recording, and transcription built for therapy. The compliance and the clinical tooling are the point.
Can I paste client information into Wren?
Yes — that's the whole idea. Every user signs a BAA with Wren Clinical, and Wren holds BAAs with its AI and transcription subprocessors, so PHI is handled compliantly.
What can I ask it?
Anything you'd ask a knowledgeable colleague or a general AI assistant: think through a case, draft a document, brainstorm interventions, reformat something, or ask a clinical or documentation question.
Does it train on my data?
No. Your conversations, transcripts, and notes are never used to train any AI model.

Explore more use cases

Transcription Telehealth documentation