#jslaybaugh - Bad Actor
1 messages ยท Page 1 of 1 (latest)
Hello! Can you provide a bit more detail? I'm not sure I fully understand the situation. How are you determining the Customer is bad after adding the card to them?
So we're getting hundreds of new accounts a day and they're all clearly spam (names like "asdfas asdfasdf", but they're just being used to scam cards and see which are valid. The customer apparently gets created and the source is stored, but then shortly afterwards, when Stripe attempts to do the subscription/invoice etc, the card gets declined because of the fingerprint block we put in place
See
How did the fingerprint block get put in place though?
Between the Customer signing up and the first charge?
no, the fingerprint is the same across all these signups
because the fingerprint is associated with the person/location/device that is doing this, from what i understand
Oh, I see what you're saying. Yeah, this is expected. Radar rules and blocklists don't apply to attaching cards directly to a Customer. If you were using Setup Intents you could enable Radar for those though.
Ok... but you're saying that is by design? Why wouldn't they all use the same validation?
It's just not how the system works. I don't know why it was designed this way. I do know that Radar is focused on prevent fraud for payments, and there's no payment involved when directly attaching card info to a Customer.
Radar also supports Setup Intents as option on top of that.
The option is in the Dashboard here: https://dashboard.stripe.com/settings/radar
Out of curiosity, in addition to things like setup intents and payment intents... are these rules applied when tokenizing a card?
No.
Note that Radar, our integrated solution for automatic fraud protection, performs best with integrations that use client-side tokenization.
that's from https://stripe.com/docs/api/tokens
Complete reference documentation for the Stripe API. Includes code snippets and examples for our Python, Java, PHP, Node.js, Go, Ruby, and .NET libraries.
I don't know what specifically those docs were referring to, but for client-side tokenization the only things that are checked are things like cvc and postal code validation - we don't check the additional radar rules at that point in time
ok thanks for confirmation!