Google AI: For the last few days, there has been a lot of discussion on social media that Google is silently training its Gemini AI model by reading your Gmail. Many viral posts are claiming that your emails, attachments and chats are being used for AI brainpower and that is why Google has made changes in its privacy policy. Some posts are even advising people to turn off smart features of Gmail. However, Google says that all this is completely wrong.
Speaking to The Verge, Google spokesperson Jenny Thomson clearly said, “All these claims are misleading. We have not changed any settings, Gmail’s smart features have been in place for many years, and we do not use Gmail content to train the Gemini AI model.” Despite this, many users have started nervously checking their Gmail settings.
IMPORTANT message for everyone using Gmail.
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 Settings menu in TWO locations.Retweet so everyone is aware. pic.twitter.com/54FKcr4jO2
—Dave Jones (@eevblog) November 19, 2025
Where does the real confusion begin?
The misunderstanding increased after an update in January. Google had given a new option in which users could control the smart features of different apps separately. That means, if you want, you can turn off smart features in Gmail, Calendar or Docs but keep them on in Maps or Wallet.
But according to the report, the settings of some users were automatically reset and the features which they had previously turned off were found to be on again. It is not clear how big this technical glitch was, but it increased people’s suspicion.
These smart features of Gmail not only check spell but also provide facilities like automatically adding flight tickets to calendar, online order tracking, auto-summary of emails, important reminders etc. When you turn these on, you’re told that your Workspace data is used only to personalize your experience. Google says that it has nothing to do with AI training.
Google’s claim
Google is repeatedly reiterating that Gmail content does not go to train Gemini 3 or any AI model. Your email only helps in providing better suggestions, reminders and improvements to you and not in AI model training.
This controversy has arisen at a time when Google’s new and powerful Gemini 3 AI model is in the headlines. Google CEO Sundar Pichai said that this model can understand information in a human way, can break down complex data layer by layer, and is better than ever at capturing the context of long conversations. Gemini 3 can process almost any type of data – handwritten notes, research papers, video lectures, recipes.
Is your Gmail secure?
The company clearly says, “Neither is AI trained by reading your emails, nor is any setting silently changed.” However, the growing concern about AI and privacy is natural, so it is not a bad idea to check your Gmail settings. This will tell you how your data is being used and who is actually reading your emails.
Also read:
The entire house will be ready in 24 hours! Spider robot did wonders, human jobs are in danger
