A financial services firm is concerned about employees inadvertently pasting sensitive customer account information into...

The requirement has two parts: govern use of generative AI destinations and inspect content for sensitive information. An AI Guardrail rule supplies application-specific control for AI interactions and can monitor or block prompts, responses, and supported file uploads. A Data Loss Prevention policy supplies the content inspection by matching configured data identifiers and classifications, then enforcing the selected action when customer-account data is detected. Cisco’s Secure Access guidance explicitly describes DLP rules that can monitor or block prompts, responses, and embedded files exchanged with selected AI applications. Blocking every HTTPS connection on port 443 or disabling all browsing would be excessively broad and would prevent legitimate business traffic. An endpoint firewall rule based on known service IP addresses neither performs content classification nor reliably covers cloud services whose addresses change. Consequently, the required pair is AI Guardrails plus DLP. Cisco Secure Access DLP guidance