LoFP LoFP / authorized testing, demonstrations, or approved workflows may allow responses that match pii or pci guardrails. review the application owner, policy, response context, and whether the data was expected to be disclosed.

Techniques

Sample rules

Cisco AI Defense Allowed Sensitive Response

Description

Detects Cisco AI Defense response events that were allowed and matched a privacy guardrail for personally identifiable information or payment card data. These events may indicate that sensitive information was returned to an AI application user and should be reviewed promptly.

Detection logic

`cisco_ai_defense`
event_action="allow"
event_message_type="response"


| eval guardrail_ruleset_type=mvjoin('event_policy_guardrail_assocs{}.policy_guardrail_assoc.guardrail_avail_ruleset.guardrail_ruleset_type', ", "),

       guardrail_entity_name=mvjoin('event_policy_guardrail_assocs{}.policy_guardrail_assoc.guardrail_avail_entity.guardrail_entity_name', ", "),

       application_name=if(isnotnull('genai_application.application_name') AND len(trim('genai_application.application_name'))>0, 'genai_application.application_name', "unknown"),

       connection_name=if(isnotnull('connection.connection_name') AND len(trim('connection.connection_name'))>0, 'connection.connection_name', "unknown"),

       policy_name='policy.policy_name'


| where mvfind('event_policy_guardrail_assocs{}.policy_guardrail_assoc.guardrail_avail_ruleset.guardrail_ruleset_type', "(?i)\\bPII\\b
|personally identifiable information
|\\bPCI\\b
|payment card industry")>=0


| table _time event_date event_id event_message_type event_action resource_type
        application_id application_name connection_id connection_name model.model_name
        policy_name guardrail_ruleset_type guardrail_entity_name


| `cisco_ai_defense_allowed_sensitive_response_filter`