The AWS Outage and How CEHC Prepares Students to Respond

ALBANY, N.Y. (Oct. 21, 2025) — This week’s Amazon Web Services (AWS) worldwide outage was a stark reminder of how vulnerabilities in centralized cloud infrastructure can trigger widespread digital disruption. The failure, originating in the US-EAST-1 region due to a DNS resolution issue with the DynamoDB API, cascaded through critical services like EC2, IAM and Lambda. Within hours, thousands of applications across finance, healthcare and government sectors experienced major interruptions, a textbook example of a cyber cascading failure.

This is precisely the scenario we've been preparing for at the CCR Cyber Cascade Risk Laboratory at the College of Emergency Preparedness, Homeland Security, and Cybersecurity at UAlbany. Since 2023, our team has been building a comprehensive Functional Dependency Network Analysis (FDNA) simulation that models all 33 AWS server regions. Our simulation treats each region as a node in a complex dependency network, allowing us to predict failure propagation, identify critical weak links and visualize the dynamic impact of a degradation in real-time, exactly the kind of dependencies that failed.

By modeling the weighted influence and failure thresholds between regions, it provides a quantitative framework to understand why any single point of failure (like the October 20th US-EAST-1 Region incident) has such a disproportionate impact. The simulation doesn't just show what happened; it allows us to run "what-if" scenarios to proactively identify and reinforce other potential single points of failure across the global cloud ecosystem before they cause the next outage.

We are now advancing this work by integrating federated learning, moving beyond analysis to active resilience. This next-generation approach enables distributed intelligence where regions learn locally without sharing sensitive data, adaptive mitigation for dynamic rerouting during failures, and predictive recovery to optimize restoration paths.

As cloud services become increasingly central to global digital infrastructure, the need for sophisticated risk modeling becomes critical. Today's cloud outage reminds us that our digital world rests on complex interdependencies that demand rigorous scientific analysis.

Interested in learning more about our FDNA simulation or exploring collaboration opportunities? Connect with us!
Cyber Cascade Risk (CCR) Lab | Take a Tour of the Lab! | Connect with Dr. Ariel Pinto

Dr. Ariel Pinto serves as Professor and Chair of the Department of Cybersecurity at UAlbany’s College of Emergency Preparedness, Homeland Security, and Cybersecurity (CEHC). His work bridges theory and practice to advance understanding of complex digital risks. Dedicated to mentoring future cyber leaders, Dr. Pinto is committed to transforming hard-earned lessons in risk into innovative solutions for a secure digital future.