Home / Technology / Google Denies Using Gmail for AI Training
Google Denies Using Gmail for AI Training
22 Nov
Summary
- Google disputes claims of using Gmail content to train AI models.
- The company states 'smart features' have existed for years without policy changes.
- Some users reported being opted back into smart features unintentionally.
Viral social media posts and articles have suggested Google is using personal Gmail messages and attachments to train its AI models. Google has publicly refuted these claims, asserting that no policy changes have occurred and user settings remain unaffected. A company spokesperson clarified that Gmail's 'smart features,' which enhance user experience, have been available for many years.
Despite Google's reassurances, the company has acknowledged that some users may have been inadvertently opted back into these smart features. This comes after a January update that allowed for independent management of smart features across Google Workspace and other Google products. Users are advised to review their settings to ensure they align with their preferences.
These smart features include functionalities like order tracking and automatic calendar additions for flights. While enabling these features within Google Workspace involves agreeing to personalized experiences based on content and activity, Google maintains this does not equate to using email content for AI training purposes, aiming to quell user concerns about data privacy.
