I am not sure I get the connection between AI code holding up review processes and data breaches.

The post made a pretty clear claim, I thought: the volume of apps being sent through is so extreme that they can't keep up with their review process.