Google says it isn’t using your Gmail to train AI

You may have seen the now-viral warnings that Google is snooping around your Gmail account and using your emails to train its AI models.

“IMPORTANT message for everyone using Gmail,” reads the most viral post on X from user Dave Jones. “You have been automatically OPTED IN to allow Gmail to access all your private messages & attachments to train AI models.”

This Tweet is currently unavailable. It might be loading or has been removed.

Numerous outlets have now reported on these claims, usually followed by a step-by-step process to shut off Gmail’s Smart Features, which supposedly stop Google from using your emails.

But it’s not true, according to Google. The company is not using your Gmail account to train its AI, even if users leave the Smart Features on.

“These reports are misleading – we have not changed anyone’s settings, Gmail Smart Features have existed for many years, and we do not use your Gmail content for training our Gemini AI model,” said a Google spokesperson in a response provided to Mashable. 

“Lastly, we are always transparent and clear if we make changes to our terms of service and policies,” Google continued.

Mashable was initially skeptical about the claims that Google was using users’ emails to train AI unless users opted out of a feature. Smart features, which essentially integrate Google’s AI model Gemini into a user’s Google Workspace account, have been available for some time now. This is not a new feature.

Additionally, Google has been explicit about its commitment to user privacy regarding AI models, such as Gemini. 

“Your data stays in Workspace,” reads Google’s policy page on Gemini. “We do not use your Workspace data to train or improve the underlying generative AI and large language models that power Gemini, Search, and other systems outside of Workspace without permission.”

Some users seem to have interpreted giving Gemini access to Gmail by turning Smart features on as equivalent to granting Google that permission. However, Google has previously addressed this on its AI on Google Workspace page, stating that this is not the case. The permissions a user grants with these Smart features are specifically to let Gemini access the data for that user’s use cases, not AI training.

Users are certainly justified in questioning the AI policies of all tech companies. We are aware that numerous companies have already trained their AI models on data and content for which they did not have permission to use. And any user who wants to turn off AI features to be safe is well within their right to do so.

But, this specific claim about Google training its AI using emails from Gmail accounts unless users specific turn off a few features appears to be completely false.