Google has rewritten a line in its privacy policy that long served as a firewall between its most personal products and its advertising business, adding an exception that could let data from Gmail, Drive and Google Photos inform the ads users see when they connect those apps to the company's AI features.
The change, dated October 16 and first spotted by Android Authority, was reported by The Verge and Ars Technica. The old text was categorical: "We don't show you personalized ads based on your content from Drive, Gmail, or Photos." The new version keeps that promise but attaches a condition, covering data that users choose to make available to other Google services that support personalized ads.
What makes the exception operational rather than theoretical is Personal Intelligence, the feature Google released in early 2026 that lets users hand data from Gmail, Photos and Workspace apps to its AI products. The point is better answers: an assistant that knows your hotel confirmation email can recommend somewhere to eat nearby. The same connection is what the policy now points to, and Google's own example in the document is search, which "may process some of the data that you allow Search to access to personalize content, including ads."
Google argues this is a clarification, not a change in practice. In a statement to Ars Technica, the company said connecting apps is entirely optional, that ads "will not be based directly on your Gmail inbox or Photos library," that sensitive information will not be used to tailor ads, and that advertisers never receive the underlying content or identifying details such as a name or email address. Google also says it does not serve ads based on sensitive categories such as race, sexual orientation, religion or health.
The gap between "not based directly" and "may inform" is where the substance sits. Ars concluded that the statement amounts to a confirmation that Workspace and Photos data may feed ad targeting, because the summaries and insights drawn from that data shape the answers AI Mode produces, and those answers are delivered alongside sponsored results. The Gemini app itself carries no ads today, but AI search surfaces do.
The timing is awkward for a company that has spent the year arguing that its AI features are worth trusting with personal context. Personal Intelligence is a broad grant of access, and Google promotes it with prompts and pop-ups that critics say do not fully spell out the consequences. A user who switched it on to get better restaurant suggestions has, in effect, also widened the pool of signals available to the ad system.
Users who want out have to go looking. The connected apps can be disconnected from the Personal Intelligence settings inside Gemini, or through the Google account under Data and privacy, then Personalization settings, then Manage Personalized Recommendations in Search Services, then Connected Content Apps, where Workspace and Photos connections can be turned off. Disconnecting or deleting the original content does not necessarily remove information already stored in search services history.
The larger pattern is that consent designed for one purpose keeps getting reused for another. Personal Intelligence was framed as a trade of context for better answers; the policy update quietly adds advertising to the list of things that context can touch, without a product announcement or a notification to users. Regulators in Europe and elsewhere have been focused on exactly this kind of scope creep, and a policy document is easier to point at than a settings menu buried four levels deep.
Comments (0)
Log in to join the discussion
Log InNo comments yet