Google Quietly Updates Privacy Policy to Let Gmail and Photos Data Feed Ads

Google has carved out an exception to its long-standing promise not to use Gmail, Drive, and Google Photos content for advertising, quietly revising its privacy policy so that data from those apps can help personalize ads when users connect it to the company’s AI features.

The change, which took effect Oct. 16 and was first spotted by Android Authority, adds a conditional clause to a pledge Google has made for years.

The updated policy now reads that Google doesn’t show personalized ads based on content from Drive, Gmail, or Google Photos “unless you choose to make data from these services available to other Google services that support personalized ads.”

The catch hinges on Personal Intelligence, Google’s push to make its AI tools more useful by letting features like AI Mode in Search and Gemini draw on a user’s own data.

Connecting Gmail, Drive or Photos to those tools delivers more tailored answers, such as surfacing a hotel reservation buried in an inbox. But under the revised wording, that same connected data can now also inform which ads a person sees.

Google stressed the limits. Content from the three services still isn’t used for ad targeting by default, advertisers don’t get access to the contents of anyone’s Gmail, Drive or Photos, and the company says it won’t share personally identifying details without permission.

In practice, that means Google isn’t scanning private emails and photos wholesale to sell ads, but users who opt into its AI conveniences may be trading some ad privacy for them.

Users can opt out by disconnecting the apps under Connected Content apps in their Google Account’s Data & privacy settings, or by turning off personalized recommendations, though Android Authority noted that previously collected data may linger in a user’s Search history even after disconnecting.

The quiet rollout echoes recent Google privacy posturing around its AI ambitions. As Decrypt reported in May, Google’s Chrome browser began silently installing a multi-gigabyte on-device AI model on users’ computers, and separately removed a privacy claim tied to its Gemini Nano model, both moves that drew scrutiny over how the company handles user data as it races to embed AI across its products.