That sounds a lot like the stories I heard about Paypal, and their fraud detection team. There was an article interviewing someone about it. Back in the day, Paypal had to build tools to detect fraud because no one else was doing it at scale, at the time. I recall something about investigators outside of Paypal started to use those tools as well. I would not be surprised if Stripe had to create similar tools for themselves as well. Without transparency, it's hard to say what the false positive rates are, or what has changed in internal processes or in the regulatory landscape that might have triggered something different.

Stripe abstracts away a lot of the complexities involved in the payment and banking world, but there's a ton of infrastructure there related to detecting fraud and money laundering. Unfortunately, the lack of transparency makes what might be a leaky abstraction look like a Kafkaesque bureaucratic nightmare.