Zoom Addresses Backlash, Updates Terms Regarding AI Training on Customer Data
Video conferencing platform Zoom has made revisions to its terms of service following significant backlash over recent updates that allowed for artificial intelligence (AI) training on customer data. The changes, which were implemented in March without much attention, sparked concerns regarding the company’s control over customer data for AI training purposes. In response to the criticism, Zoom released a blog post asserting that it would not engage in the activities outlined in its terms of service. As a result of continued blowback, the company subsequently updated its terms, explicitly stating that it will not train AI models on consumer video, audio, or chats without customer consent.
The controversy stemmed, in part, from Zoom’s experimental AI tools such as IQ Meeting Summary and IQ Team Chat Compose, which rely on machine learning and AI algorithms. While account owners must provide consent to use these tools, additional participants are only presented with the option to accept the terms or leave the meeting. This lack of an opt-out provision raised concerns about user privacy and consent, as the terms appeared to grant Zoom extensive rights over customer content, including the ability to redistribute, extract, reproduce, and process it in various ways.
Alex Ivanovs, writing for Stack Diary, highlighted the alarming nature of this data usage, emphasizing the potential for significant debate around the issue. However, Zoom’s Chief Product Officer Smita Hashim clarified in the recent blog post that consent is required from account owners and administrators before sharing data for AI training, which is exclusively intended to enhance the performance and accuracy of AI services. Hashim further emphasized that customer content remains under their ownership and control, even if Zoom uses it for value-added services like livestreaming.
To address concerns, Zoom’s updated terms now explicitly state that customer consent is necessary for AI training using audio, video, or chat content. The revised terms assert that Zoom will not use customer content, including education records or protected health information, to train its AI models without consent.
Hashim’s blog post emphasized that Zoom aims to empower account owners and administrators with control over AI features and decisions. The company seeks to provide transparency regarding its data usage policies and how they may affect specific customer groups.
In conclusion, Zoom has responded to the backlash by taking steps to clarify and restrict AI training on customer data. With the updated terms now in place, the company aims to address privacy concerns and ensure that customer consent remains a priority in the use of their content.
