ModelCharter

Is Claude safe for work?

Low risk · 20

Anthropic · AI assistants · facts (high-confidence)

Claude is low-risk for default at-work use (20/100): it trains on your data unless you opt out, and holds SOC 2 Type II.

Watch out: Even when a consumer opts out of training, conversations flagged for safety review can still be retained up to two years and used to improve models without notifying the user.

Data and compliance facts

Trains on consumer-tier data
Opt-out
Trains on business-tier data
No
Training opt-out available
Yes
SOC 2 Type II
Yes
ISO 27001
Yes
ISO 42001 (AI management)
Yes
GDPR Data Processing Addendum
Yes
HIPAA BAA
Yes
EU data residency
Unverified
SSO / SAML
Yes
Data retention
Consumer: 30 days if training off, up to 5 years if training on; API inputs/outputs deleted after ~7 days and never used for training; safety-flagged content may be retained up to 2 years regardless of opt-out.
Safer tier
Claude Enterprise (Claude for Work)

Why it scores 20 out of 100

  • +14Trains on your data unless you opt out. Training is on by default on the consumer tier; you must find and toggle the opt-out.
  • +6No EU data residency. Data cannot be guaranteed to stay in the EU.