While it’s clear the government has made progress since the initial guidance was issued, there’s still much to be done to support overall safe federal AI.