Not sure what the real difference or benefit is here, but my phone app (google, but not pixel) already labels calls as "likely fraud" or "likely spam".
You would have to answer the call for this to be useful, no?
It's for when you're already on the phone with them, if they start asking for your name, address and social security number the systems says "scam detected".
I imagine they have a MASSIVE database of numbers that make calls in certain patterns. If a new number makes 10,000 calls that had been dormant, it gets flagged.
I will only be impressed when Google forwards said scam caller to AI Alzheimer's Granny and wastes all their time. Only then will I applaud. It is past time to fight back.