Red Flags in AI Privacy Policies: What to Look For
When reviewing privacy policies or terms of use for AI tools, keep your eye out for these common risk signals:
❌ Red Flag #1: “We may use your data to improve our models.”
→ This often means your input may be stored and reviewed.
❌ Red Flag #2: “We do not accept responsibility for incorrect outputs.”
→ You need strong disclaimers and manual review protocols.
❌ Red Flag #3: “We do not allow use for sensitive information, including health data.”
→ This disqualifies the tool for any PHI use.
❌ Red Flag #4: No mention of HIPAA, BAA, or compliance frameworks
→ If they serve healthcare, they should at least mention security standards.
❌ Red Flag #5: “You agree to indemnify us...”
→ Look out for legal clauses that shift all liability to you, the user.
🟩 Green Flags to Look For:
HIPAA or SOC 2 Type II mentioned
Ability to turn off training
Option to enter into a BAA
Enterprise plan with enhanced security
