@britechguy It does matter who typed it in, because the person typing in the data accepts the risks associated with using the platform.
If I as a merchant type information into Square, or Wells Fargo, or anything I use... and those systems are breached the impacted owners of that data's legal recourse is to sue me for the damages.
If I configure a system such that the user has to input data into the aforementioned service to pay me, while they're typing in all that stuff they have to accept the terms and conditions of using the service themselves. The process is branded for that service, not mine. So if there's a breach, they now have legal recourse to sue the vendor in question directly. Because they interacted with that vendor, at my behest perhaps, but they still made the interaction.
Who inputs the data into what branded fields is what makes the liability determination in a court room. You can clearly see this in the PCI compliance questionnaires, and the banks track who's at what level based on the level of certification requested during the compliance process.
By the way this is worst case thinking stuff, if something goes wrong typically the user's card is changed and they aren't on the hook for any actual losses. But if they were on the hook, it's better if they can sue the big boy directly, instead of having to sue the vendor they actually paid, and have that vendor sue the merchant provider.