That's probably the biggest reason to have good in-house security people. They don't have a financial interest to make breaches or lie about them. It's in their best interest to keep everything secure, and continue to look for new ways to attempt breaking into their own stuff.
I've never felt good about letting third parties in to do security testing. When someone above my rank decided to let a 3rd party do external tests, they'll pick anything and make it sound disastrous. One place was bitching about anything.
They complained that we had the current version of Bind running on the DNS servers. "But people can do DNS requests!" Yup.
They flagged the fact that we dropped unwanted traffic at the firewall. Yup. Get over it. They were upset it took forever to scan the network. Good.
They flagged us for having a web server providing static content. They were upset they couldn't find any way to exploit CGIs or do SQL injection. Yup. That was kind of the idea.
There were a whole bunch of other trivial things that they flagged us for. Then they were brought to the office, and got upset that we didn't provide wifi. Nope, that's a security risk. They wanted to plug their laptop into our network, so they were only given external access. Again, they bitched. But letting an unknown computer owned by an unauthorized party plug into our network is a security risk.
They eventually gave up trying to bully us into dropping our security precautions and gave us a pass.
I already habitually ran tests with privileged access to make sure even if all layers of protection failed, nothing really bad could happen.
Honestly, if they are given everything, they can find something. Give them administrative rights to everything, and credentials to everything, they can find something. Like, email accounts can be accessed with full admin rights. Funny how that works.