Google users are being alerted to a quiet change that may have allowed the company to scan private emails and attachments for artificial intelligence training without clear user approval.
The issue centers on a setting that was switched on automatically, raising fresh concerns about privacy, consent, and how personal data is handled in the race to build powerful AI tools like Gemini.
The issue came to light after electronics design engineer Dave Jones from Australia revealed that Google had enabled a feature in October 2025 that allows email content to be used to improve AI systems.
According to Jones, the change applied to both personal and work emails and did not require users to actively agree to it, meaning many people may not have known their data was being accessed.
Jones warned users that turning off the feature is not simple and requires changes in two separate places within Google settings. Desktop users must first disable “Turn on smart features in Gmail, Chat, and Meet” under the Smart Features section. They must then go further and opt out again under “Manage Workspace smart feature settings” before saving the changes.
However, users attempting to protect their privacy quickly discovered an unexpected drawback. Once the opt-out process is completed, inbox categories such as promotions, social, and updates disappear.
This leaves users with a single unfiltered inbox that may contain thousands of mixed emails, making it far harder to manage daily messages.
As a result, many users feel forced to choose between keeping their inbox organized and limiting how their data is used. Critics argue this trade-off unfairly pressures people to accept scanning in order to keep basic email functions.
The backlash has already reached the courts. In November 2025, a class-action lawsuit was filed in the United States by Illinois resident Thomas Thele. The case accuses Google of secretly using Gmail users’ email history without proper consent.
The lawsuit covers all Gmail users in the country whose data was scanned by Gemini AI after the feature was turned on.
The concern is not limited to desktop users. People accessing Gmail through Android or iOS apps are also affected. As long as users are logged into their Google accounts, their emails and attachments remain subject to scanning under the same settings.
Many users are worried about the type of information stored in their inboxes. Emails often contain bank details, medical information, legal records, and private conversations. Allowing this data to be used for AI training has raised fears about misuse or exposure if systems are breached.
Google has pointed to its privacy policy to defend its approach, stating, “Google uses information to improve our services and to develop new products, features and technologies that benefit our users and the public.”
The company has also said it does not directly use Gmail content to train Gemini. Still, critics argue that automatic opt-in systems weaken trust and leave users unsure about how their data is truly handled.
For smartphone users who want to opt out, the steps are similar. They must open the Gmail app, tap the menu icon, go to Settings, select their account, then choose Data privacy. From there, users need to turn off Smart features and personalization and also disable options under Google Workspace smart feature settings.
Opting out comes with further limits. Users lose tools such as email auto-complete, spelling suggestions, and the ability to quickly add calendar events from messages. Google allows users to reverse the decision at any time if the loss of convenience becomes too much.
Tech analysts note that even opting out does not fully solve the problem. Privacy can still be affected if the other people involved in an email conversation keep the feature turned on. Others have voiced doubts that Google will keep the settings disabled long term, fearing the feature could be turned back on later without clear notice.
As users weigh convenience against control over their personal data, the debate over AI training, consent, and user trust continues to grow. With legal action underway and scrutiny increasing, the issue highlights the growing tension between advancing technology and the right to privacy.
