AWS Cost Explorer Bug Triggers Trillion-Dollar Billing Alarms
A critical bug in AWS Cost Explorer erroneously displayed trillion-dollar billing estimates to customers, causing widespread alarm despite actual charges remaining unaffected.
Amazon Web Services (AWS) customers experienced significant alarm on July 16-17, 2026, when the AWS Billing and Cost Management Console and its Cost Explorer tool began showing astronomically high projected cloud costs. Reports emerged of estimated monthly bills reaching into the trillions of dollars, which triggered automated budget alerts and raised immediate concerns about unauthorized AWS usage or system misconfigurations.
AWS acknowledged the issue via its official Health Dashboard and its support account on X (formerly Twitter), assuring users that the inflated figures were confined to estimated billing data. The company emphasized that actual charges and verified usage records were not impacted by the anomaly. The problem began around 7:38 PM PDT on July 16, with engineering teams investigating the cause of the inaccurate estimated billing information.
By 3:03 AM PDT on July 17, AWS identified the root cause as an issue within the unit pricing calculations of its estimated billing computation subsystem. "The displayed billing estimates do not reflect actual usage and charges," AWS stated, clarifying that the error lay in the projection layer rather than the core metering or invoicing systems. This means that legitimate usage data was being multiplied by incorrect unit prices, leading to the implausible financial projections.
AWS Cost Explorer is a vital tool for customers to monitor historical usage, forecast future spending, and identify cost-saving opportunities. During an active billing period, it relies on estimated pricing, which is distinct from the final invoices generated from validated metered service usage. The bug affected this estimation process, creating a disconnect between projected and actual costs.
While AWS stated that no immediate customer action was required, the company warned that a full recovery would take several hours post-mitigation, as it needed to recompute estimated billing data across all affected accounts. The incident sparked considerable discussion on social media, with users sharing screenshots of their alarming estimates and expressing relief that their actual costs were unaffected.
This event also highlighted the sensitivity of automated AWS Budgets alerts. While the estimates were erroneous, the incident serves as a reminder that unexpected billing spikes can indeed signal genuine issues such as compromised credentials, exposed access keys, cryptomining operations, misconfigured resources, or accidental deployment of resource-intensive workloads.
AWS advised customers to continue monitoring their accounts and to independently validate usage through tools like CloudTrail, AWS Config, and Cost and Usage Reports. Security and FinOps teams were urged to confirm that the anomalous alerts subsided after AWS completed its recalculation process. If cost increases persist after the AWS incident is resolved, customers are advised to investigate for unauthorized activity or configuration changes.
AWS has not provided a definitive timeline for complete resolution but committed to issuing further updates as mitigation efforts progressed. The incident underscores the importance of robust monitoring and validation processes, even when relying on cloud provider tools for cost management.