Has AI changed how your team verifies high-risk requests?

With AI-generated voice and video becoming much more convincing, it feels like relying on someone's voice or appearance is becoming less reliable for sensitive requests.

I'm curious whether this has changed how your team handles things like password resets, wire transfer requests, account changes, or privileged access approvals.

Have you updated your verification process because of AI-assisted impersonation, or are your existing procedures still working well?

I'd be interested to hear what's actually been effective in real-world environments.