Is Google digging through your Gmail chats for AI training? Company breaks silence on viral buzz

Rumours have surged across social media in recent days, suggesting that Google has been quietly tapping into Gmail accounts to train its Gemini AI models. According to widely shared posts, the company is allegedly scraping emails and attachments and has even altered its privacy policy without informing users. These claims insist that users must disable Gmail’s “smart features”, such as autocorrect and predictive suggestions, to prevent their data being used for AI development.

What was circulating online?

A YouTube influencer triggered much of the discussion after sharing an “urgent” message, warning users that Gmail had automatically opted them into allowing access to their private messages and attachments for AI training.

The post shared on X (formerly Twitter), accompanied by multiple screenshots, urged people to manually disable Smart Features in two separate locations within 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 Setting menu in TWO locations,” the post read.

Google’s clarification

Google issued a public response on Saturday, November 22, using its official Gmail account on X to counter the accusations. The company stated that the claims circulating online were inaccurate.

“Let's set the record straight on recent misleading reports. Here are the facts: We have not changed anyone’s settings. Gmail Smart Features have existed for many years. We do not use your Gmail content to train our Gemini AI model,” Google wrote.

The company further emphasised, “We do not use your Gmail content to train our Gemini AI model. We are always transparent and clear if we make changes to our terms or policies.”

How users responded online

Reactions on social media ranged from supportive to sceptical. One user commented, “Gmail, it’s just people hating on you guys! That’s all. You guys are on with what I call A.G.E.” Another claimed, “Only saying this so they don't get sued in the EU.”

Others questioned the specifics of Google’s statement. One user wrote, “The assurance is noted, but the language matters. The statement confirms Gmail content is not used to train Gemini. The real question is whether usage metadata (things like send times, recipients, and topic frequency) is utilized for general AI feature development or profiling. That data stream is often the more valuable and less explicitly covered component in these assurances.”

Another sceptical comment read,““to train our Gemini AI model” This sounds like Google-speak for "We use your emails to train an AI model, just not Gemini". That is disheartening.”

Another user remarked, “If not Gemini, what AI models do you train with the data?”

One person asked, “So why is the feature on by default. Why is all features used for tracking users Data on by default. And why is it that even Apps are being installed on Android Phones without consent of the User. And Google have had many Antitrust Lawsuits and been found GUILTY!”