Google is denying recent claims that it uses Gmail messages and attachments to train its AI models. Viral social media posts and news articles suggested users must turn off smart features to prevent their emails from being used for AI training. Google says these reports are wrong.
Google Responds to Training Claims
According to The Verge, Google spokesperson Jenny Thomson stated the company has not changed user settings. She said Gmail smart features have existed for many years. Google does not use Gmail content to train its Gemini AI model.
The viral posts claimed Google changed its policy. They said the only way to opt out was by disabling smart features like spell checking. One article from Malwarebytes made similar claims. Google calls these reports misleading.
However, one Verge staff member found they had been opted back in to smart features after previously opting out. In January, Google updated its smart feature personalization settings. The update lets users turn off features for Google Workspace and other Google products independently.
What Smart Features Actually Do
Gmail smart features include more than spell checking. They enable tracking orders and adding flights from Gmail to your calendar. The settings page says users agree to let Google Workspace use their content and activity to personalize their experience across Workspace.
Settings You Can Control
Google maintains this does not mean handing over email content for AI training. The company says it processes data to provide smart features and improve them. But this processing stays separate from AI model training.
Users can still check their settings if they want to review their choices. The smart features can be turned on or off in Gmail settings. Google says no recent policy change has altered how it handles Gmail data for AI purposes.
