Skip to main content
AI Governance
Rule Watch™

Vendor Privacy Scorecard · Anthropic

Claude: is it safe for client information?

A strong writing and analysis assistant. Commercial plans keep your data out of training; personal plans ask each user to choose.

Depends on plan and settings

Bottom line

Safe for client work on Claude Team or Enterprise. On personal Free, Pro or Max plans, each user controls whether chats help train models, so a firm cannot prove the setting.

Does it train on your data? By plan

  • Free, Pro and Max (personal): User or admin setting

    Each user chooses in privacy settings whether chats are used to train models.

  • Team and Enterprise: No training

    Commercial terms: not used for training.

  • API: No training

    Not used for training. Zero data retention is available by agreement.

What an auditor will ask

Trains AI on your data? Depends on plan
Commercial plans and the API do not train. Personal plans depend on each user's setting.
Retention you can control? Depends on plan
Enterprise offers custom retention. Zero data retention is available on the API by agreement.
Signs a HIPAA BAA? Depends on plan
Anthropic signs a BAA for qualifying Enterprise and API customers through sales.
Staff may read your content? Depends on plan
Content flagged for safety or policy violations may be reviewed.

Security certifications

  • SOC 2 Type II
  • ISO/IEC 27001
  • ISO/IEC 42001

Set it up the safe way

  1. Use Claude Team or Enterprise for work involving client information.
  2. Ban personal Claude accounts for work in your Staff AI Acceptable Use Policy.
  3. Sign a BAA before any health information goes in.
  4. Save the commercial terms and your review date.

Anthropic trust centerBusiness privacy terms

Summarized from Anthropic's published terms, last checked 10/1/2026. Vendors change terms. Confirm every point against the agreement you actually signed. This is not legal advice.

See every vendor scorecard

Using Claude at work? An unvetted tool can cost you your license.

When an employee pastes client information or confidential files into Claude without the correct safeguards, licensing boards and regulatory bodies hold you responsible. AI Governance Rule Watch™ documents your approved tool roster, vendor reviews, and staff guidelines in permanent diligence records ready for any audit.

Protect my business for $67/month Or $670/year · 2 months free

Founding member rate. Cancel anytime. Plus applicable taxes. Renews until canceled. Not legal advice and not a compliance certification.