Zoom Says It’s Not Stealing Customer Content to Train AI Models

variety.com:

Todd Spangler NY Digital Editor A recent change to Zoom’s terms of service pertaining to AI set off alarm bells in Hollywood and the tech world — with some interpreting the videoconferencing provider’s update as granting the company royalty-free rights in perpetuity for customer video calls and presentations for the purposes of training AI models.

In response, Zoom said it doesn’t use any customer audio, video or chat content for training AI without consent. In addition, the company says, the “service-generated data” it collects is to “make sure that we aren’t unwittingly being used to spam or defraud participants” using its AI-powered features.

A post Sunday on the blog StackDiary said Zoom’s new terms of service appeared to look as if the company was “willing to be all its chips on reusing other people’s content for AI training.” The post cited section 10.4 of Zoom’s TOS, which says, “You agree to grant and hereby grant Zoom a perpetual, worldwide, non-exclusive, royalty-free, sublicensable, and transferable license and all other rights required or necessary to redistribute, publish, import, access, use, store, transmit, review, disclose, preserve, extract, modify, reproduce, share, use, display, copy, distribute, translate, transcribe, create derivative works, and process Customer Content and to perform all acts with respect to the Customer Content,” including for the purpose of “machine learning” and “artificial intelligence” for the “improvement of the services, software, or Zoom’s other products, services, and software.” “This effectively allows Zoom to train its AI on customer content without providing an opt-out option, a decision that is likely to spark significant debate about user privacy and consent,”.