Google has denied viral claims that Gmail content is being used to train its Gemini AI, calling the reports misleading and inaccurate.
Quick Summary (TLDR):
- Google says it does not use Gmail messages or attachments to train its Gemini AI model
- Viral social media posts and a Malwarebytes article sparked widespread concern over Gmail privacy settings
- Smart Features like spell check and calendar syncing are not linked to AI model training, according to Google
- Users are encouraged to double-check settings due to a January update that may have re-enabled features
What Happened?
Google is responding to a surge of viral misinformation that alleges the company is using Gmail data to train its Gemini AI models. The claims suggest that users must disable Gmail’s Smart Features, including spell check, to opt out of AI data sharing. Google has firmly rejected these assertions.
According to company spokesperson Jenny Thomson, Google has not made any changes to user settings or privacy policies related to Gmail and AI. The Smart Features have been part of Gmail for years and serve only to personalize the user experience, not to feed data into AI training.
Google Pushes Back Against AI Training Rumors
Concerns were sparked by a Malwarebytes article and viral social media posts that accused Google of quietly updating its policies to allow Gmail messages and attachments to be used in AI training. The article claimed that opting out of this required disabling Smart Features.
IMPORTANT message for everyone using Gmail.
— Dave Jones (@eevblog) November 19, 2025
You have been automatically OPTED IN to allow Gmail to access all your private messages & attachments to train AI models.
You have to manually turn off Smart Features in the Setting menu in TWO locations.
Retweet so every is aware. pic.twitter.com/54FKcr4jO2
Google was quick to refute the allegations. In a statement to The Verge, spokesperson Jenny Thomson said, “We do not use your Gmail content for training our Gemini AI model.” She emphasized that Gmail’s Smart Features, such as order tracking and automatic calendar entries, are meant to enhance convenience but have no role in AI model development.
Google insists that no recent changes have been made to its Gmail settings or AI training practices. The company believes the current misunderstanding stems from an update made in January 2025, which allowed users to separately manage Smart Features for Google Workspace and other services like Maps and Wallet. This may have led to some features being unintentionally re-enabled for certain users.
What Are Gmail Smart Features?
Smart Features in Gmail are designed to make routine email tasks simpler. When enabled, they can:
- Check spelling and grammar automatically as you type
- Track packages and orders directly within emails
- Detect flights and add them to your Google Calendar
While these features do analyze some user data to function, Google says that data is only used to personalize your experience within Workspace, not to train its Gemini AI or any other machine learning models.
The company explains on its settings page that enabling Smart Features means you “agree to let Google Workspace use your Workspace content and activity to personalize your experience across Workspace.” It stops short of allowing that data to be harvested for external AI training.
Why Users Should Double-Check Gmail Settings
Despite Google’s clarification, there is still room for confusion. One Verge journalist noted that although they had previously disabled Smart Features, they later found them turned back on. This highlights the importance of regularly reviewing privacy and personalization settings, especially following any platform updates.
Google maintains that no Gmail content is being used for AI training, but acknowledges that the settings update in January could have reactivated Smart Features without clear notification.
For users who value tight control over their data, it is a good idea to revisit the Gmail personalization and Workspace settings and adjust them as needed.
Daily Research News Takeaway
Honestly, this situation is a great example of how fast misinformation can spread online. I get why people were worried. When you hear that your personal emails might be feeding an AI model, it’s natural to panic. But in this case, the panic seems to be based on a misunderstanding. Google has been crystal clear: your Gmail is not training Gemini AI. That said, if privacy matters to you, go ahead and review those settings. A quick check never hurts, especially after a system update.
