Link: https://www.wsj.com/articles/fraudsters-use-ai-to-mimic-ceos-voice-in-unusual-cybercrime-case-11567157402

I think this is an important issue to discuss in our community. There are a number of community members that have publicly released their speech synthesis models and they have cloned other people’s voices without consent. I believe that sets a dangerous precedent.

It’s important that speech synthesis models are not posted online for everyone to use. Those open source models may easily be abused.

It’s important that we don’t clone celebrities voices without their explicit consent. A speech synthesis model of a celebrity can put that person at great risk of identity fraud.

What do you all think?

I am not without my own biases. I do work at a speech synthesis company. We have a concrete focus on the ethics of speech synthesis.