Cybersecurity · VIDEO
Human-approved actions by AI on cardholder data
The PCI Security Standards Council has published the Security Considerations for AI Systems, a document that provides guidelines on data protection in AI sys...
KEY TAKEAWAYS
- Clear definition of purposes, permissions, and accesses for every AI system
- Mandatory human approval for actions on cardholder data and management of shadow AI
- Preventive and continuous security tests to prevent bypassing of restrictions
- Protection of sensitive data and credentials with separate management and encryption
- Rigorous evaluation of external vendors and specific incident response plans
