Zoom Adopts Customer Data for AI Training under Updated Terms of Service
Zoom, the prominent video platform, has strategically updated its terms of service to encompass its authority to leverage specific customer data for advancing its AI capabilities. The revision, which came into effect on July 27, brings forth an intriguing facet of Zoom’s AI strategy by permitting the use of customer data for training and refining its artificial intelligence and machine-learning models.
When a user enables Zoom’s generative AI features, released in June, the company has them sign a consent form allowing Zoom to train its AI models using their customer content.
Utilizing Service-Generated Data for AI Advancement
Zoom’s recently modified terms of service prioritize sections related to software licensing, beta services, and compliance. However, a closer examination of the fine print reveals a pivotal development in the company’s AI journey. The updated terms now grant Zoom the capacity to employ certain segments of customer data, encompassing data on product usage, telemetry, diagnostic information, and comparable data amassed by the organization, to enrich its AI capabilities. Notably, there is no provision for opting out of this data utilization.
Navigating AI Training Amidst Ongoing Debate
This move by Zoom comes amidst an ongoing public discourse concerning the appropriate extent to which AI services should draw from individuals’ data, irrespective of its level of aggregation or anonymization. Noteworthy AI entities, including OpenAI’s ChatGPT, Google’s Bard, and Microsoft’s Bing, as well as image-generation tools like Midjourney and Stable Diffusion, rely extensively on a vast array of internet text or images for their training. However, this practice has sparked legal challenges from creators who assert that their work is being mirrored in AI-generated outputs.
The Consent Paradigm
Zoom’s revamped terms explicitly outline the parameters within which it can access, utilize, collect, modify, share, and process “Service Generated Data” for various purposes, including the advancement of machine learning and artificial intelligence. The company emphasizes that it refrains from utilizing audio, video, or chat content for AI training without explicit customer consent, underscoring the significance of user approval in these processes.
In other words, once a user gives consent to use the new AI features, Zoom will be able to use the data to further enhance its AI.
Empowering AI through Customer Content
In a notable stride toward enhancing its AI capabilities, Zoom introduced two novel generative AI features in June. These features—a meeting summary tool and a chat message composition tool—are available on a trial basis to customers, affording them the discretion to decide their usage. Upon activation, users are required to consent to the usage of their individual customer content for training Zoom’s AI models. The company assures users that their content serves the sole purpose of augmenting the precision and effectiveness of AI-driven services.
Zoom’s Perspective and User Empowerment
A spokesperson for Zoom conveyed that customers wield the authority to enable generative AI functionalities and independently determine whether to share customer content with Zoom for the betterment of its products. This approach emphasizes user agency in contributing to the evolution of Zoom’s AI capabilities.
In conclusion, Zoom’s updated terms of service underscore the company’s intention to leverage select customer data for AI model training and enhancement. While sparking reflection within the broader AI discourse, this strategic move aligns with Zoom’s commitment to user consent and empowerment, as the company embarks on its AI-driven journey.