As a DevOps engineer for a rapidly growing social media company, you're part of the monitoring team for their flagship web application. The application, deployed on Managed Instance Groups behind an HTTP(S) load balancer, is generating logs at a rate that exceeds the project's logging API quota. Despite implementing exclusion filters in Cloud Logging, the issue continues. What could be the underlying problem?