It’s a question of defensibility if some regulator came to you and asked why you allowed certain individuals through because a black box told you without any additional context.

And I don’t think anyone builds their own KYC from ground up - it’s more whether you have context for a decision with your own vendor implementation vs a black box yes or no.

Which in turn means you benefit from everyone else who's already been stung. KYC solutions' risk:reward ratios start high, but asymptotically approach 0 as the provider learns from its customers' problems. If you're signing up for a well-established one, you're able to free-ride on all the learning that's already been done.

You can have your cake and eat it, too: - Start with a third-party KYC solution - Slowly develop your own and compare the results with the one from the provider - When your own solution provides something similar to the other one, you can abort the subscription