Is AI Used to Audit AI?

To what extent are AI used to audit AI? And if these could not create a sort of loophole whereby, you know, who audits the audit algorithm and so on. What sort of safeguards can be put in this regard? Well, for sure, many kinds of audits are going to deploy AI to audit AI. So how do we understand that method if the AI that's used to conduct the audit is itself a black box? It's where standards become so important so that it's clear what the method of auditing is.