I have worked on systems that had to do this in the long-ago past, and the headline is simply wrong. There are many 8-bit EBCDIC code pages that support many l

I have worked on systems that had to do this in the long-ago past, and the headline is simply wrong. There are many 8-bit EBCDIC code pages that support many languages (multi-hundred page book of them last I looked at it), at least as many as ISO-8859 alphabet soup encodings... and there are also shift-in/shift-out double byte EBCDIC encodings for Japanese, Korean, and Simplified and Traditional Chinese. They've been around since the 1980s at least. Since nesrly the beginning of utf8 acceptance, there is also UTF8-EBCDIC (shudder) The bank developers just didn't write their code in a way that can support the multi-codepage way of storing EBCDIC.