Zoom revises service terms so it could train AI on user data

Zoom has updated its terms of service to establish a “perpetual” right to use customer content to test and train AI but later clarified it will not use video, audio, or chats without consent.

The initial revision, dated July 27th, required users to agree to their data potentially being utilized by the company for the purposes of “training and tuning” artificial intelligence (AI) and machine learning (ML) algorithms and models if they wish to continue making video calls on the platform.

Following public scrutiny, Zoom once again updated its terms of service on Monday (August 7th) to clarify that it "will not use audio, video, or chat customer content to train our artificial intelligence models without your consent."

The updated service terms still require users to “hereby grant Zoom a perpetual, worldwide, non-exclusive, royalty-free, sublicensable, and transferable license and all other rights” to the customer content.

These include rights 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.

The document no longer specifically mentions "AI and ML training" among the acts Zoom would be entitled to perform in relation to customer content but more vaguely refers to "service development, marketing, analytics, quality assurance, machine learning, artificial intelligence, training, testing, improvement of the services, software, or Zoom's other products, services, and software, or any combination thereof."

The policy also describes service generated data, which includes telemetry data, production usage data, diagnostic data, and similar data that isn’t considered customer content as “solely retained” by Zoom.

The user is required to consent to the company’s access, use, collection, creation, modification, distributions, processing, sharing, maintenance, and storage of service generated data “for any purpose” as long as laws permit it.

The change in the terms of service was first observed by Substack Diary, which said it “effectively allows Zoom to train its AI on customer content without providing an opt-out option.”

In response, Zoom's spokesperson said that "customers decide whether to enable generative AI features, and separately whether to share customer content with Zoom for product improvement services."

Zoom was founded in 2011 and experienced rapid growth during the pandemic. The company decided to reduce its workforce by 15% earlier this year, ending the boom period, but is pinning its hopes on AI as the future driver of growth, in what is a reflection of general trends in the industry.

More from Cybernews:

Five most common cybersecurity vulnerabilities in 2023

Apple removes, then restores popular Russian podcast after backlash

Apple removes, then restores popular Russian podcast after backlash

Moscow civil servant and politicians’ addresses leaked say pro-Ukrainian attackers

Colorado education department admits data breach

Subscribe to our newsletter

Leave a Reply

Your email address will not be published. Required fields are markedmarked