Google Disputes Claims It Trains AI on Gmail Data Without Explicit Opt-Out
MOUNTAIN VIEW, CA – Google is pushing back against assertions that it utilizes user Gmail data to train its artificial intelligence models unless users actively disable specific features. The company maintains it has been transparent about its AI training practices adn that user data within Google Workspace remains protected.
The controversy stemmed from concerns that Google’s “Smart features,” integrating its Gemini AI model into google Workspace, might be leveraging user emails for AI development without clear consent. However, google asserts this is inaccurate, clarifying that access granted through smart features is solely for individual user experiences, not for broader AI training purposes. This debate arrives as scrutiny of tech companies’ AI data practices intensifies, following lawsuits alleging unauthorized use of copyrighted material to train AI models.
Google’s policy page on Gemini explicitly states, “Your data stays in Workspace. We do not use your Workspace data to train or improve the underlying generative AI and large language models that power Gemini, Search, and other systems outside of Workspace without permission.” The company further addressed the misunderstanding on its AI on Google Workspace page, emphasizing that enabling Smart features does not equate to granting permission for AI training.
The clarification comes amid a broader conversation about data privacy and AI development. Numerous companies have faced legal challenges for training AI on data they did not have the rights to use, including a recent settlement with Anthropic over alleged copyright infringement. Google encourages users who are concerned to disable AI features, affirming their right to do so.
