The topic of how artificial intelligence will be used is a veritable minefield and on Monday, Zoom stepped in it.
The video conferencing company, which became a household name during the pandemic, adjusted its terms of service to give the company the right to use “service-generated data” to train its AI systems.
That led to a bit of mass hysteria (at least online) and concerns that Zoom was taking our video calls and using them to train their AI. Zoom responded with an updated blog post denying this.
“To reiterate: we do not use audio, video, or chat content for training our models without customer consent,” Chief Product Officer Smita Hashim said in an updated blog post (in bold typeface).
Similar language also appears on the updated terms of service after the controversy erupted. On Friday, Zoom went a step further.
“Following feedback, Zoom made the decision to update its Terms of Service to reflect 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,” the company said, noting that it has also updated its terms of service to make it clearer.
You can forgive the hysteria given the explosion of popularity of AI and the speed at which they’ve developed. ChatGPT, in particular, has served as a catalyst for virtually every major tech company to invest in their own AI systems, despite warnings from experts who fear the unforeseen repercussions.
AI also feeds into the broader concern that your personal information and ideas are being fed into the systems to make them smarter, often without your consent or proper compensation.
Much of the concern is tied to two AI-powered features introduced in June by Zoom, IQ Meeting Summary and IQ Team Chat Compose. Both launched in June, with Meeting Summary automatically generating a write-up of the meeting based on what was said and its own AI systems, and Team Chat Compose automatically writing responses for you.
The owner of the meeting has to enable the features, so there’s an opt-in element to it, and they can always disable data sharing.
But those who attend the meeting don’t have any say in whether they share their info if the administrator has it on. They’ll only get a message at the beginning disclosing that the feature is begin used.
The company had previously said it is leaving it up to customers to determine whether to enable the AI features.
“Zoom customers decide whether to enable generative AI features, and separately whether to share customer content with Zoom for product improvement purposes,” a company spokesperson said on Monday (The company points to the new, more explicit statement issued Friday). “We’ve updated our terms of service to further confirm that we will not use audio, video, or chat customer content to train our artificial intelligence models without your consent.”