Learn
Home Foundations Glossary Research
Do
Goal Bank Prompts Tools Tasks
Adapt
Domains Settings Patterns
Verify
Reading LLM Writing Case Studies Policies AI Updates Resources

Claude HIPAA Coverage Is Enterprise-Only — Team Plans Don't Count

Anthropic offers BAA coverage through its API and a HIPAA-ready Enterprise plan. Free, Pro, Max, and Team plans are not covered — a common and consequential misunderstanding.

July 7, 2026 caution claude baapolicyprivacy
Clinical Impact: Only an organization-signed HIPAA-ready Enterprise plan (or an API deployment under a BAA) is eligible for PHI workflows. Personal, Pro, Max, and Team accounts are not — no matter how your workplace plan is labeled.

What’s True Right Now (verified July 2026)

Anthropic offers Business Associate Agreements (BAAs) on two surfaces: the first-party API, and a HIPAA-ready Enterprise plan. The Enterprise path is sales-assisted — an organization’s primary owner accepts a click-through BAA and explicitly enables HIPAA mode in organization settings. Coverage doesn’t happen by default, even on Enterprise.

Not covered, per Anthropic’s own documentation: Claude Free, Pro, Max, and Team plans, Cowork, the developer Workbench/Console, and features in beta.

Why This Matters for SLPs

The plan names invite a dangerous assumption. “Team” sounds organizational, and organizational sounds compliant. It isn’t:

  • Private practice SLPs on Team plans: your plan cannot be covered by a BAA. Entering PHI is a HIPAA violation regardless of how professional the account tier sounds.
  • Hospital and health system SLPs: your organization needs the HIPAA-ready Enterprise plan with HIPAA mode actually enabled — not just “an Enterprise contract.” BAAs signed before December 2, 2025 covered API usage only, so even organizations with older agreements may not be covered for the chat product.
  • School-based SLPs: FERPA (not HIPAA) governs most student records, and a BAA doesn’t address FERPA at all. Districts providing school-based health services may carry dual obligations — that’s a compliance-officer conversation, not a plan-tier checkbox.

Important Limitations Even With Coverage

  • HIPAA mode is explicit. The org owner must accept the BAA and enable it. Being on Enterprise is not enough — verify, don’t assume.
  • Covered models require 30-day data retention and aren’t available with zero-data-retention configurations.
  • A BAA doesn’t make everything safe. You still follow your organization’s AI policies, review every output, and de-identify where you can. Defense in depth.

What to Do

  • If your workplace tells you “we have Claude, you can use it for documentation,” ask specifically: is it the HIPAA-ready Enterprise plan, is HIPAA mode enabled, and does our BAA postdate December 2025?
  • If you’re on a personal, Pro, Max, or Team plan: keep PHI out entirely. De-identification rules from PHI Safety apply in full.
  • If you’re evaluating for an organization, start from Anthropic’s own coverage pages rather than vendor summaries — plan coverage has changed over time and will again.

Sources

SLP/IO Assistant

Powered by Claude · No PHI accepted
AI assistant for clinical workflow support. Never enter student names, DOBs, or identifiable information.
Hi! I'm the SLP/IO assistant, an opinionated AI grounded in clinical practice. I can help with goal wording, note structure, ethical reflection, and navigating LLMs responsibly. What are you working on?