Theoretically AI could reduce delays in approvals because a machine can approve someone for cover faster than a human. But what happens with the denied people? The article does not seem to mention whether refusals still have an appeal option which would then guarantee human intervention (one presumes).

In principle, more humans should be freed up to work on appeals if AI is used for the initial decision. But knowing the Trump regime, those freed up people will have a job security problem.

In Europe, the GDPR would theoretically¹ protect people from this. Automated decision making is generally banned but has exceptions. But even when automated decision making is legal, there is a legal obligation to have the possibility of human intervention.

Some US states have their own dilluted GDPR variant, but I think none of them give a shit about automated decision making.

¹ I say “theoretically” because GDPR enforcement is a disaster.