Yeah the graph for that one looks exactly the same shape. I wonder if they were depending on some azure component somehow, or maybe there were things hosted on both and the azure failure made enough things failover to AWS that AWS couldn't cope? If that was the case I'd expect to see something similar with GCP too though.
Edit: nope looks like there's actually a spike on GCP as well
It's possibly more likely that people mis-attribute the cause of an outage to the wrong providers when they use downdetector.
Definitely also a strong possibility. I wish I had paid more attention during the AWS one earlier to see what other things looked like on there at the time.