>Look there are limits to what I'm able to talk about, but consider that a fraud subsystem functioning in transaction-time is running much more slowly than some

>Look there are limits to what I'm able to talk about, but consider that a fraud subsystem functioning in transaction-time is running much more slowly than something running in request time. I was suggesting checking one day snapshots of access counts by IP to see what the distribution looks like to then determine what to block. 2fa: I know that chase asks me for 2fa every time I change IP, and I don't think there's a way to stop that. Should be easy for them to require 2fa on every login for any IP with a high rate of requests.