Skip to main content
🚨Early AccessYour AI Team Just Got Smarter. Unlock Fable 5.1, GPT-6 Astra & Gemini Flash 3.8 with Team or Business.
Help Center / Security & Admin / Setting your Data Loss Prevention (DLP) policy
Security & Admin

Setting your Data Loss Prevention (DLP) policy

Updated · 3 min read

DLP scans chat messages before they leave your organization for an AI model, and decides what happens to four classes of sensitive data:

CategoryCovers
PIINames, emails, phone numbers, addresses, SSNs
PHIMedical records, diagnoses, medications, providers, lab results
Financial PIICredit cards, routing and account numbers, IBAN, tax IDs
SecretsAPI keys, private keys, JWTs, connection strings, passwords

Setting the policy

From Settings → Security, choose one of three actions for each category:

  • Allow — the message goes through unchanged
  • Redact — matched values are replaced with a placeholder before the message is sent, then it goes through
  • Block — the message doesn't send at all, and the sender sees why
This is set once per organization, not per workspace or session — it applies to every member, every chat, org-wide. Only organization owners and admins can change it; everyone else sees it read-only.

Related articles

→Setting PII redaction→Setting workspace policies

Was this article helpful?