Zoom backtracks again on its AI policy, says it won't use video calls ...

Zoom, the video-conferencing platform, wanted to train its

AI

tool with customer data. But then it said it would not do so without the user’s consent. Now,

Zoom

has backtracked, saying it will not use any user data to train its AI models or of third parties.

Previously, Zoom, in its terms and conditions said that it would not train its AI tools “without customer consent.” In the blog post published earlier,

Smita Hashim

- Chief, that account owners and admins can activate certain features in the trial phase, which allow the company to use user data for AI training. This raised concerns about customer consent since it was unclear how users could consent when the meeting organiser had already enabled the feature and agreed to the terms.
In section 10.4 of Zoom's Terms of Service, customers grant Zoom a perpetual, worldwide, non-exclusive, royalty-free, sublicensable, and transferable licence to use their data for product and service development, machine learning, and artificial intelligence. However, after public outcry, a new clause was added stating that Zoom will not use audio, video, or chat customer content to train its artificial intelligence models without the customer's consent.

Smita Hashim, Chief Product Officer at Zoom, said that after receiving feedback regarding Zoom's updated terms of service. The blog clarified that Zoom does not use any audio, video, chat, screen-sharing, attachments, or other customer content to train its AI models.
Now, Zoom has updated its terms of service again, writing that its generative AI tools do not use "communications-like" customer data to train AI models for Zoom or third-party use. “Zoom does not use any of your audio, video, chat, screen sharing, attachments or other communications-like Customer Content (such as poll results, whiteboard and reactions) to train Zoom or third-party artificial intelligence models,” reads the updated terms of service.