This article is from Aug of this year, but I just found it now & to be honest, I’m quite concerned:
Zoom has updated its terms of service to include the use of certain customer data for training its artificial intelligence (AI) models.
Zoom’s terms clearly state that they can access, use, collect, create, modify, distribute, process, share, maintain, and store service-generated data for various purposes, including AI training and tuning. However, customer content like messages, files, and documents are not included in this category. Zoom has clarified in a blog post that they do not use audio, video, or chat content for training their models without customer consent.
However, I’,m not sure I believe them. How would we know if they would or would not?
A friend the other day told me that she says nothing when first answering a call to her cellphone from an unrecognized telephone number. She lets the “caller” speak first. If she hears nothing and the “caller” eventually hangs up, then she assumes that it was not a call from a human. When I asked her why she does this, she said that she does not want her voice recorded for use later in AI-generated speech – e.g., in a call to a relative to say something such as, “I am stuck in London and lost all my money and credit cards. Can you send me some money by Western Union?”
I do the same but for a different reason, I let the caller speak first because I believe that once an automated voice hears you speak it knows that it’s a valid number. But your friend’s take on it is interesting.