Why do you think this is an issue?

The "text to speech" feature is a tool, just like an instrument.  Unless you are making a clone of the software and selling it as your own, there's no issue.

Otherwise, as long as the "text" is not someone else's copyrighted material, what problem could there be?