What is a common mistake with AWS availability and disaster recovery in AWS?

Declaring a system highly available because it uses multiple AZs while ignoring regional failure, data recovery, dependencies, or failover testing gives false confidence.