Two major AWS outages in October and November 2025 exposed dangerous flaws in cloud infrastructure, crippling services from Amazon.com to Coinbase and costing an estimated $581 million. The root cause? A latent race condition in DynamoDB DNS and a flawed order system — highlighting systemic risks of centralized cloud dominance.