It is intentional, I'm just saying it's ridiculous. the Google way is to treat everything like a programming problem. That works decently well for bits, but really breaks down when working with humans.
The thing is that it doesn't break down except in the extreme cases, and even then it's often less bad than the edge cases with traditional human-mediated customer service. Consider the Wells Fargo accounts snafu, or the Comcast customer nickname issue. Or PayPal just in general. Is Google actually worse, or are they the kind of better that makes their rarer screwups seem worse. Consider in your evaluation just how many people they deal with using their automated systems.
Extreme cases like using a popular new credit card to pay? Please. Google uses an algorithm to ban people for life without recourse, it's insane. Not everything can be reduced to an algo, sometimes you need to make exceptions. It can even make your algo better (thousands of people are going to try using their new Apple cards in the coming weeks).
As others have mentioned, we only currently have one data point on this. However...
Google is well known for terrible customer support. And if you think of it, it makes sense. After all, if you can server 100M users via an automated support systems and only 0.01% have issues a year, it's probably far, far cheaper to just drop those "problem" users, than hire thousands of customer support persons.
Note: just threw the above numbers out, but you get the gist.
EDIT:
Speaking of algorithms, I wonder if the OP was below some adsense fiscal threshold of 'it's worth it to intercede here with a level 4 manager' or some blather.
You don’t have the data to conclude that it doesn’t work well. We could very well be commenting on an exception out of millions of perfectly executed automated fraud detections.
