If the bank still relies on legacy software and IT standards, well it's just its fault. They cannot expect people with diacritics or other non-ASCII characters

If the bank still relies on legacy software and IT standards, well it's just its fault. They cannot expect people with diacritics or other non-ASCII characters in their name just to spell it incorrectly because their systems do not support Unicode in the twenty-twenties. Maybe their IT team had other priorities than replacing EBCDIC with Unicode (or whatever they find more appropriate for their systems), but this is an indicator of poor interest in technological progress by the bank itself. It reminds me some banks that gave millions to Microsoft to keep ATMs running Windows XP after its end of life. Edit: I elaborated a bit more and I realized that it might be more difficult than just replace the character encoding standard to a more modern one. For example, the name of the account owner likely needs to match exactly the holder name on the credit card associated with the account, and I'm not sure if diacritics can be embossed correctly on the card.