I was thinking that character doesn't adequately describe software-based vocalists. Most "characters" are roles voiced by actors, but artists like the famed Miku Hatsune don't fit this mold. So I think a new artist type is required.

I don't know what to call it though. Here are a few possibilities:

• voice synthesizer - good catch all term, we can put Fred (the most popular artificial voice), Siri & Cortana here too.
• artificial performer - also accurate
• software performer - I'm making sure to emphasize performance as these are not creators (yet).