The Unicode encoding to save space group similar characters from the Chinese, Japanese and Korean into a single byte sequence (Han Unification) and use other hints to decide between each character, but on Anki Desktop the Chinese variant is almost always chosen, leading to words with wrongly written kanji.
I’ve been looking for a solution that doesn’t involve modifying global variables (like in Inconsistent unicode normalization by epistularum here on the forum), when i remenbered that Anki is just a web viewer and allows HTML tags, which leads to the solution just being wrapping the Japanese text in <p lang="ja">{{Front}}</p>.
is this setting saved on the deck or is it a device setting? because if it saves on the deck, then a font that is available on both the Desktop App and Mobile would be needed, using <p lang="jp"> at least is device/font agnostic
Currently it’s not possible to specify more than one font. I don’t know how AnkiDroid handles this; if it’s like iOS where you can specify an order for language preferences, maybe adding Japanese to the list above Chinese would help.