Handling faces in source footage before adding voiceover?

Curious if anyone here has a workflow for dealing with faces in raw footage before bringing it into Genny — specifically for clips that include bystanders or people who didn't consent to being on camera.

I've been running footage through an automated blur step first, then adding the AI voiceover afterward. Works well as a pre-processing stage. Wondering if others have solved this differently or built something similar into their pipeline.