Shark Tank investor and consumer-products expert Lori Greiner is urging people who use the world’s biggest email platform to check a little-known setting she says could expose sensitive messages to AI scanning.
Greiner claims many of Gmail’s roughly 1.8 billion users don’t realize certain inbox features can allow AI-powered tools to process email content, potentially including highly personal details.
The 56-year-old “Queen of QVC” said: ‘Google doesn’t want you to know this, but they’ve been allowing AI to scan every single one of your emails,’ adding that this can include ‘financial documents, tax information and personal conversations’.
She also said the fix is quick—taking “in like 30 seconds”—as long as you know where the option is located.
According to Greiner, users should open Gmail, click the gear icon at the top right, then choose “see all settings.” From there, she advises scrolling to “smart features and personalization.”
Next, she says users can find the option labeled “Turn on smart features in Gmail, Chat and Meet” and switch it off.
To further reduce AI-driven processing across Google services, she recommends checking the section for other Google products and locating additional Smart Features controls.
In that area, she says to verify that Smart features in Google Workplace and Smart features in other Google Products are also turned off.
The Daily Mail reports Greiner adding: “Once you’ve done that, just click save changes, and you’re all set. Your emails are now protected from AI scanning, and Google cannot access your personal information anymore.”
Even as Greiner tells users to take the situation seriously, Google has pushed back on the implication that these tools are secretly harvesting data, explaining what the features are designed to do.
Gmail can use Google’s AI assistant, Gemini, to power functions such as summarizing email threads and suggesting “smart” replies intended to speed up routine responses.
Google says that while these tools may process message content to complete a user-requested task, any personal data involved isn’t used to train the AI system. The company also maintains that Gemini only accesses what’s needed to fulfill a request.
It further claims the data associated with those requests isn’t retained, and that the processing happens within the user’s Gmail experience rather than being broadly fed into other Google systems.
Still, not everyone is satisfied with that reassurance, and some users argue there are downsides to turning the features off.
One Gmail user wrote on Reddit: “FYI disabling it also disables spell/grammar checking and autocorrect. You know – the non-AI stuff GMail has had for decades. Simply Google’s way of forcing everyone to do what they’re told.”