Why TTS if it's just a static phrase? More likely that a random developer recorded that sentence and the wav is hardcoded in all the units.

It might be that the law around voice recordings are not as simple as we might imagine.