Let's think about the logic of the requirement. AS9100 Rev C says:
7.5.2 The organization shall validate any processes for production and service provision where the resulting output cannot be verified by subsequent monitoring or measurement…
Some process results are easy to measure after they've been done. If a cutting process is done, and the characteristic I'm interested in is length, I could use a caliper or micrometer to measure the length. This doesn't sound like a special process.
Some process results are harder to measure. If a bonding process is done, and I'm interested in the strength/integrity of the bond, I may have a hard time checking it. I may need to do destructive testing on a sample to validate that the process worked as intended. This sounds like a special process. I can't do 100% destructive testing and still ship product.

It makes sense that we should validate this process so that we're confident that the pieces we didn't test were done properly and have the strength/integrity that we need.
Now let's look at ultrasonic and radiographic testing. What's the intended result? A test report of
the results of another process? Can the resulting output "be verified by subsequent monitoring or measurement?" Yes. I have a test report. By the way, is testing even a process for production? Does it change the part in any way, or does it just
verify another process? This doesn't sound like a special process to me. If this is a special process, then the caliper measurement from the first scenario is too, because there is nothing I can measure on the product that will tell me if I've measured it before.
(I'm putting on my flame-proof underwear now.) Maybe I'm crazy, but since radiographic testing and ultrasonic testing are actually monitoring/measurement processes themselves,
I don't believe that they're "processes for production" as referenced in AS9100C 7.5.2. As such
I don't think they have to be classified as special processes. If a field failure occurred in a weldment due to a failed process, the root cause of the failure would never be "the test process wasn't done properly." The root cause would be "the weld process wasn't done properly." Many people point to the NADCAP list of "special processes" to show that UT and RT should be treated as special processes. We should consider that PRI financially benefits from having more processes fall under their purview and their NADCAP accreditations.
Having said that, I have been an AS9100 auditor for years. RT and UT are both important verification processes and proper controls should be in place to ensure that they're done correctly. Many times during an audit I've found that
the client has identified RT or UT as special processes. In these cases I've audited them as special processes.