- Newest
- Most votes
- Most comments
I believe two things you can follow now:
- Filter by Event Name If you're seeing false alerts, consider filtering or suppressing alerts for:
"eventSource": "freetier.amazonaws.com",
"eventName": "GetAccountPlanState"
- Tag Events as Internal Use your SIEM or alerting tool to tag these events as internal AWS service calls, even if the IP looks external. You can base this on the userAgent or consistent structure of the event.
Hey,
Hope you're keeping well.
The GetAccountPlanState calls you’re seeing are triggered automatically by the AWS Console or CLI when checking Free Tier usage, and they’re served via AWS edge infrastructure. In some cases, CloudTrail logs the public IP of the CloudFront edge location handling the request rather than an internal AWS service hostname, which is why you’re seeing IPs from different geographies. This isn’t an indication of unauthorized access, but it can cause false positives in geo-based alerts. If you need to filter these, you can match on eventSource=freetier.amazonaws.com and eventName=GetAccountPlanState in CloudTrail or your SIEM to exclude them from alerts.
Thanks and regards,
Taz
Hi Taz, thanks for your answer. Are you aware of any other services where this behavior happens? As far as I can tell, Free Tier events are the only ones which show random IPs in the sourceIP field. Since these alerts have started to appear, I have already set rules to ignore them in SIEM, but this behavior seems unique to the Free Tier service. I asked this question to see if there is any reason why this service behaves differently in the Cloudtrail logs, and to point out that this unique behavior for the Free Tier service creates unnecessary noise, and if other services did the same, it would be a bigger problem.
Relevant content
- asked a year ago
- asked 2 years ago
- AWS OFFICIALUpdated a year ago
- AWS OFFICIALUpdated 2 months ago

Thanks for the response Kidd Ip. This is the first time I've ever seen events logged like this which is why I raised the question, it seemed very odd. Considering there are not any other AWS Services or API calls which exhibit this behavior where random IPs are logged for the sourceIPAddress in Cloudtrail events, I'm not sure if something is wrong with the account settings that is causing this, and I can not tell why this would have started to happen all of a sudden. Of course these Freetier read only events are not very critical, but if other events (such as s3:DeleteBucket) were to show the same behavior, we would not be able to ignore these events in our SIEM. This behavior would diminish the usefulness of Cloudtrail if the data in the events can not be trusted, and would need to be ignored.