Zoom Denies A.I. Fears Sparked by Recent Update

Zoom has denied that the latest update to its terms of service will allow the video conferencing software company to utilize user data to train artificial intelligence (AI).

The notion had been propagated online after the changes were noticed by a freelance software developer. Social media users decried what they saw as an invasion of privacy by a platform that has an estimated 800 million customers worldwide.

However, the company said the update was to "further clarify" that it would not employ user data to train AI without consent.

The changes come at a time when attitudes to AI are a mix of excitement about the technological advancements it is already bringing and concerns about the ethical questions it poses.

Experts have cautioned that generative AI—which is used to mimic human content—has the potential to replace human workers and repurpose intellectual property. Some have also expressed concern about Zoom claiming ownership over things discussed in video calls that users or companies retain sole control over. Newsweek approached Zoom via email for further comment on Tuesday.

Zoom has said that the AI-related updates made clear that it would only use users' personal data with their consent, and that use of content users produce would only be used for training if they opt in to using its new AI features.

On Monday, the company updated its terms of service regarding customer content—e.g. videos, files, and transcripts—and what it describes as service-generated data, such as telemetry and diagnostics data.

Section 10.2 of the revised terms of service reads: "You consent to Zoom's access, use, collection, creation, modification, distribution, processing, sharing, maintenance, and storage of Service Generated Data" including for the purpose of "machine learning or artificial intelligence (including for the purposes of training and tuning of algorithms and models)."

Alarm was raised when this clause was coupled with another, which would grant the company "perpetual, worldwide, non-exclusive, royalty-free, sublicensable, and transferable license and all other rights required or necessary" to use customer content for, among other things, "machine learning, artificial intelligence, training, testing, improvement."

Experts warned of a risk to user privacy in allowing data to be used to train AI software, while Alex Ivanovs, who first wrote about the changes, said it "effectively allows Zoom to train its AI on customer content without providing an opt-out option."

"Time to retire Zoom, who is basically wants to use/abuse you to train their AI," Gabriella Coleman wrote on X, formerly Twitter. "Not using Zoom again until these over-reaching permissions are gone," Justine Bateman said.

However, after the changes gained attention online, the company responded to concerned users to say that the changes were "to further clarify that we do not use audio, video, or chat customer content to train our artificial intelligence models without your consent."

The company published an explanation of the changes by Smita Hashim, the company's chief product officer, who stated in bold font: "For AI, we do not use audio, video, or chat content for training our models without customer consent."

"We wanted to be transparent that we consider this to be our data so that we can use service generated data to make the user experience better for everyone on our platform," she said of the language relating to service generated data in 10.2, citing an example of knowing where and when Zoom usage is greatest helping manage its data centers.

Hashim drew a distinction between the company's use of service generated data, which the new terms allow to be used for AI training, and the use of customer content, which she said was so the company could offer additional services "without questions of usage rights."

She said that the technology firm had recently introduced two AI features to their platform which provide meeting summaries and chat composition on a free trial basis—but stressed that users who chose to enable the features "will also be presented with a transparent consent process for training our AI models using your customer content.

"Your content is used solely to improve the performance and accuracy of these AI services," Hashim wrote, noting that customers in the healthcare and education sectors in particular would not have their customer content used without consent.

However, following the company's response, Ivanovs has maintained that the wording of the new terms of service gives Zoom a wider scope of usage of consumers' personal data than the company says it will use in practice.

"Unless these are rectified and clarified, this means that Zoom can do exactly what they say they can do because you are agreeing to grant all the abovementioned permissions," he wrote. "You do not have a choice to opt-out because opting-out is not part of the terms."