Technical reports have warned Gmail users about a hidden feature in their accounts that grants Google the ability to access and review message content as part of its AI technology development policies, sparking widespread concern and criticism.
It has been pointed out that all Gmail accounts were automatically selected to allow Google to scan messages and attachments for the purpose of training AI models like Gemini, without obtaining explicit consent from users.
It was explained that the activation of this feature began in October 2025, meaning personal and work emails may be read and used without their owners’ knowledge. This has led to a class-action lawsuit being filed against Google, accusing the company of secretly exploiting user data for AI purposes.
To protect their privacy, users need to manually disable “Smart Features,” which requires adjusting two different options within their account settings.
For Gmail users on computers, they must go to “View all settings,” then disable the option “Turn on Smart Features in Gmail, Chat, and Meet” within the “Smart Features” section. Following this, they need to go to “Manage Smart Feature settings in Workspace” and revoke permissions.
However, disabling this feature stops the automatic categorization of emails and removes the “Promotions,” “Social,” and “Updates” tabs, turning the inbox into an unorganized list of messages. The inbox can only be reorganized by re-enabling Smart Features, which has angered many users.
For smartphone users, “Smart Features” can be disabled by going to “Settings,” then “Data Privacy,” and turning off the “Smart Features and Personalization” option, in addition to disabling Google Workspace settings.
Users have expressed their dissatisfaction, considering that this policy makes email management more complicated, especially since it includes all accounts whether on computers or phones, increasing risks to sensitive data such as financial information and health records.
According to its privacy policy, Google has stated that it uses data to improve services and develop new products and technologies that benefit users, emphasizing that it does not directly use Gmail content to train the Gemini system. However, critics argue that the complex consent settings raise questions about the actual protection of data.
