User travel was impossible in AWS CloudTrail IAM log

VIEW RULE IN DATADOG VIEW SIGNALS

Event Summary

@userIdentity.accessKeyId: {{@userIdentity.accessKeyId}} had activity from {{@impossible_travel.triggering_locations.first_location.city}}, {{@impossible_travel.triggering_locations.first_location.country}} and {{@impossible_travel.triggering_locations.second_location.city}}, {{@impossible_travel.triggering_locations.second_location.country}} which are approximately {{@impossible_travel.triggering_locations.travel_distance}}km apart within {{@impossible_travel.triggering_locations.travel_time_human_readable}}. This indicates a potential impossible travel.

Goal

Detect an Impossible Travel event when a @userIdentity.type: {{@userIdentity.type}} uses an AWS IAM access key in CloudTrail logs.

Strategy

The Impossible Travel detection type’s algorithm compares the GeoIP data of the last log and the current log to determine if the IAM user with @userIdentity.session_name: {{@userIdentity.session_name}} traveled more than 500km at over 1,000km/hr and used an AWS IAM access key in CloudTrail logs.

Triage and response

1. Determine if the @userIdentity.accessKeyId: {{@userIdentity.accessKeyId}} for @userIdentity.session_name: {{@userIdentity.session_name}} should be used from {{@impossible_travel.triggering_locations.first_location.city}}, {{@impossible_travel.triggering_locations.first_location.country}} and {{@impossible_travel.triggering_locations.second_location.city}}, {{@impossible_travel.triggering_locations.second_location.country}}.
2. If the IAM user should not be used from {{@impossible_travel.triggering_locations.first_location.city}}, {{@impossible_travel.triggering_locations.first_location.country}} and {{@impossible_travel.triggering_locations.second_location.city}}, {{@impossible_travel.triggering_locations.second_location.country}}, then consider isolating the account and reset credentials.
3. Audit any user actions that may have occurred after the illegitimate login.