On October 16, Google revised a long-standing privacy assurance that had plainly said it would not personalize ads using content from Drive, Gmail or Photos. The new wording creates an exception when users choose to make material from those services available to other Google products that support personalized advertising.¹

The change matters because Google’s Personal Intelligence feature links data across apps to AI products such as Gemini and AI-powered Search. Introduced in early 2026, the tool promises richer answers and recommendations by letting Google draw on a user’s existing files, email and photos. But that same cross-service connection has become the fault line: a feature pitched as useful context can also become advertising context.²

Google’s policy says Search may process data a user permits it to access “to personalize content, including ads.”¹ The company insists this is not an invitation to mine an inbox or photo library directly. Connecting apps is “entirely opt-in,” it said, and ads will remain relevant to a query, the response’s context and a user’s interests — not “based directly on your Gmail inbox or Photos library.”²

That distinction is unlikely to settle the privacy argument. Critics argue the practical result is still clear: if AI Search summarizes or draws insights from connected personal data, that information can influence the environment in which ads are selected. Google says advertisers do not receive users’ files or personally identifying details, and says sensitive categories including health, religion and sexual orientation are excluded from ad targeting.¹

For users who want no such crossover, the remedy is administrative rather than automatic: disconnect Workspace and Photos from Search’s Personal Intelligence or turn off personalized recommendations in Google’s settings.¹ The broader lesson is starker: as Google makes its services talk to one another, the old firewall around personal content is increasingly defined by which switches users manage to find.