You could provide a demo with a fixed prompt that will (for example) always read the first paragraph of the Wikipedia article about avocados.

Honestly, how much of a hinderance is that? A person could just supply a recording of another person, couldn't they?

Yes, it's pretty easy to play back a video as the input text and then generate a reasonable fascimile. Here's Trump: https://sndup.net/zkbg/

but it looks like you provide the source code here? https://github.com/coqui-ai/tts. How much of a hindrance are you hoping to add?

Discourages low-hanging, hit-and-run usage that's likely to get their site shut down.

If someone wants to fake a statement there are already 100 ways to do it. Not making their servers the ones doing the deed puts a meaningful barrier in place for more casual misuse. And for serious cases like impersonation on a large scale, the resources are there to likely do better than this instant feedback model can.

Noble, but the genie will be out of that bottle soon enough, if not already.