Zoom Clarifies: No Audio or Video Used for AI Training Without Consent
Video-conferencing platform Zoom has responded to concerns regarding its updated terms of service by clarifying that it will not use audio, video, or chat content for AI training without customer consent. The controversy arose due to revisions made to Zoom’s terms in March, which seemingly granted the company the right to train AI models using call data and content. Section 10.2 and 10.4 caused particular alarm, as they allowed Zoom to use diagnostics data and customer-generated content for purposes including machine learning and artificial intelligence training. However, a new paragraph has been added to section 10.4, asserting that Zoom will not use customer content for training AI models without explicit consent.
Zoom’s Chief Product Officer, Smita Hashim, emphasized that the updates to the terms were intended to increase transparency and not to claim ownership of user data. Section 10.2 covers aggregate data on how customers use the platform, including telemetry and diagnostic data, which Zoom considers to be their own information. The intention is to leverage this data to enhance the user experience. Hashim clarified that section 10.4 aimed to allow Zoom to provide value-added services, such as meeting recordings, without concerns over usage rights. She stressed that customer consent is required for any AI training involving audio, video, or chat content. Additionally, Hashim confirmed that even if users choose to share their data, it will not be used to train third-party AI models.
Despite the lack of publicity around the terms update, Zoom has openly discussed plans to introduce AI-powered features comparable to other communication platforms. For instance, a blog post in February highlighted smart recording capabilities relying on natural language processing to summarize conversations and identify key points. Nevertheless, the intricate language and length of terms of service documents can cause customer misunderstandings or a sense of concern. Some legal experts argue for concise summaries and machine-readable formats to facilitate analysis.
While Congress introduced a bill in January that proposed one-page summaries of terms of service, including machine-readable full texts, it has not made progress yet. Zoom’s recent clarification seeks to allay any fears among users, emphasizing the importance of customer consent and assuring that no AI training will occur without explicit permission. The goal is to maintain transparency while safeguarding user data privacy.
In summary, Zoom has responded to the public’s concerns about its terms of service by clarifying that it will not use audio, video, or chat content for AI training without customer consent. The revised terms emphasize the importance of transparency and intend to allay fears about the platform’s usage of customer data. By reinforcing the need for explicit permission, Zoom seeks to build trust and reassure users about the handling of their personal information.