Zoom Lands In Trouble Over Use Of Customer Data For Training AI Models

After a recent update enabled Zoom to gather data for AI training without client approval, the popular video conferencing platform has now changed its terms of service agreement.

The update, which took effect on July 27, specifies that Zoom can access, use, and process “service-generated data” for any purpose, including machine learning or artificial intelligence.

Information on product usage, telemetry and diagnostic data, and other data gathered by the company include the service-generated data that Zoom can exploit.

It excludes consumer content like messages, files, and documents, which Zoom claims it never uses for AI training without getting permission from the customer.

In a blog post, Zoom's Chief Product Officer, Smita Hashim clarified that Zoom has updated section 10.4 of its agreement to make it clear that it "will not use audio, video, or chat customer content to train AI models without consent".

Zoom has released two new generative AI features—a tool for creating chat messages and a tool for meeting summaries—on a free trial basis for users, who can choose whether or not to use them.

However, before a user can access these services, Zoom requires them to sign a consent form authorising Zoom to use their individual customer content to train its AI models.