Back to AI Tools
Claude Cowork
Risk Overview
Moderate Risk
4.3
/10
Workflow Automation
·
Conversational AI
·
AI Agents
·
Top Risks
Discretionary deletion on account closure. $100 liability cap. Thumbs up/down feedback irrevocably training-eligible. Agentic tool use can access files and execute code.
Mitigations
Opt out of training. Avoid thumbs up/down for max privacy. Export data before closing account.
Get started
Is your team using
Claude Cowork
safely?
Highest Risk Categories
The highest-scoring risk categories identified in this assessment.
Data Retention
May at our option' delete on termination — discretionary, not guaranteed. No stated timeline. Inactive accounts may be terminated.
6
/10
Training Use
Default opt-in. Opt-out available. Hard exceptions: feedback interactions and safety-flagged content always training-eligible.
5
/10
Legal Terms
No mandatory arbitration. No class action waiver. CA courts. $100 liability cap. Broad user indemnification.
5
/10
Unlock the complete assessment
Run a free assessment to access the full AI tool risk analysis.
Updated July 2026 • AI-assisted research • Reviewed by Verax • Report an issue
Reference Documentation
Official documentation and policies referenced as part of this assessment.
Terms of Service
https://anthropic.com/legal/consumer-terms
Privacy Policy
https://anthropic.com/legal/privacy