Google responds to claims Gmail can read emails and attachments to train AI that left users terrified

Many were left horrified when reports claimed that Gmail is scanning your emails and attachments to train AI, but Google has since responded to the accusations, insisting that this isn't happening.

The rapid development of artificial intelligence has to be powered by something, and while America is increasingly seeing data centers pop up across the nation, the AI tools themselves are growing 'smarter' due to the data that they're being trained on.

Software from many of the world's biggest AI companies pull from data sources in the millions and billions, but people are growing increasingly concerned that their own data and information is being harvested and used to teach these tools.

One of the biggest 'scandals' in recent weeks came from a report that claimed Google was using any emails and attachments you send via Gmail to train its Gemini AI software, and due to the sensitive and personal nature of many email communications, this was met with an uproar from social media.

Not only was there an issue that the data was being used to train AI but that also Google was doing it without necessarily telling its user base, but the tech behemoth has now responded following these claims, denying charges that were brought up in a class action lawsuit.

Google has responded to claims that it's using Gmail data to train Gemini AI (Jakub Porzycki/NurPhoto via Getty Images)

As reported by ZDNET, a spokesperson for Google has called these new reports "misleading," adding that they "have not changed anyone's setting.

"Gmail Smart Features have existed for many years, and we do not use your Gmail content for training our Gemini AI model. Lastly, we are always transparent and clear if we make any changes to our terms of service and policies."

Malwarebytes, who shared the initial claim, has since updated its blog post after realizing they "contributed to a perfect storm of misunderstanding," adding that they claim that Gmail content was being used to train AI now "doesn't appear to be the case."

It is very much the case that Google scans your any content you send or receive via Gmail, but the express purpose of this is for things like spam filtering, categorization, and various writing suggestions, which is completely normal for an email provider.

It does appear that these settings are automatically enabled though – and there are even some cases there people have turned them off in the past even to realize that they're now reactivated following updates from Google.

Some have claimed that Google's smart features have automatically turned themselves back on (Klaudia Radecka/NurPhoto via Getty Images)

It might be worth checking if this has affected you recently, or you might simply want to deactivate them full stop, and you can do so by following these steps:

• Open Gmail on Desktop and press the cog icon in the top right to open the settings
• Select the 'Smart Features' setting in the 'General' section
• Turn off the 'Turn on smart features in Gmail, Chat, and Meet'
• Find the Google Workplace smart features section and opt to manage the smart feature settings
• Switch off 'Smart features in Google Workspace' and 'Smart features in other Google products'

There does remain still a lawsuit filed on November 11 in San Jose, California, that alleges Google 'secretly' granted its Gemini AI access to the private conversations of those using services like Gmail, Chat, and Meet.

Central to this is a supposed violation of of the California Invasion of Privacy Act, but if Google's recent clarifications hold true then these allegations would hypothetically not hold weight in court.