Google Policy Update Opens Gmail Data To Ad Targeting

Google privacy policy change removes Gmail ad targeting promise

Google quietly rewrote a key chunk of its privacy policy back on October 16, scrubbing a long-standing promise that it wouldn't touch content from Gmail, Drive, or Photos for ad personalization. Android Authority spotted the change first. The new language swaps out a flat prohibition for wording that leaves the door wide open: data from those services can now inform ad targeting whenever users share info across Google's ecosystem. The timing isn't accidental-it lines right up with the launch of Personal Intelligence, the feature rolled out earlier this year that lets people connect data from multiple Google apps to AI tools like the Gemini chatbot and AI Mode search.

The old policy was unambiguous: Google does not show personalized ads based on content from Drive, Gmail, or Photos. That assurance is gone. The updated version says ad personalization *may* draw on data from those apps when it's shared with other Google services. That equivocation matters because Personal Intelligence is built specifically to move data between services. Opt in, and your emails, documents, and photos can feed directly into AI-generated answers and search results. Since AI Mode search can surface sponsored content, the policy now admits the personal context flowing through those tools can shape the ads you see.

Google calls policy rewrite clarification not practice change

Google insists this is just a clarification, not a shift in practice. A spokesperson told Ars Technica that connecting apps stays entirely opt-in and that ads remain relevant to queries, response context, and inferred interests-not directly based on your inbox or photo library. They also say sensitive info won't be used to tailor ads, that they don't scan entire inboxes to build ad profiles, and that they don't share this data with third parties. Still, the new wording effectively confirms the reality: if personal data travels through an AI search tool, it becomes part of the signal determining which ads appear.

Developers see porous boundary between user data and ad inventory

For developers and anyone building on Google's AI stack, the update signals that the boundary between user data and advertising inventory is more porous than the company's earlier guarantees suggested. Apps integrating with Gemini or AI Mode search should assume any user content routed through Personal Intelligence might influence the ad ecosystem, even if Google filters sensitive categories. That has real implications for compliance, user trust, and how you design privacy controls in products relying on Google's models.

The pattern mirrors how Google handles AI training data. The company says it doesn't use personal data to train models, yet it uses inputs and outputs from AI interactions for exactly that. With Personal Intelligence linking Gemini to Gmail, Drive, and Photos, those outputs can contain fragments of private info. Google claims it tries to filter that content from training sets but doesn't promise perfect success. The same logic now applies to ad targeting: data shared with AI search can inform ads, and the privacy policy has been updated to reflect that reality.

Disable Personal Intelligence to keep Gmail content out of ads

Users who want to keep personal content out of the ad pipeline need to disable Personal Intelligence. The setting lives in the Gemini app under Personal Intelligence controls, where connected apps like Workspace and Google Photos can be disconnected. It's also reachable through the main Google account dashboard-Data & privacy, then Personalization settings, then Manage Personalized Recommendations in Search Services, and finally Connected Content Apps. The menus are buried several layers deep, and Google has used prompts and pop-ups that encourage enabling the feature without spelling out the advertising implications.

The broader takeaway is that Google's AI strategy depends on breaking down silos between its services. The company views its vast trove of user data as a competitive advantage for generative AI, and Personal Intelligence is the mechanism making that data available to its models. Every new connection increases the surface area for privacy exposure. Builders and users alike should treat any opt-in that moves personal content into AI-driven surfaces as a potential vector for ad targeting, regardless of how the company frames the distinction between "based directly on" and "informed by."

Frequently asked questions

When did Google change its privacy policy to allow Gmail, Drive, and Photos data for ad targeting?

Google updated its privacy policy on October 16, removing a long-standing promise that it would not use content from Gmail, Drive, or Photos for ad personalization. The new language states data from those services may inform ad targeting when shared across Google's ecosystem.

What is Personal Intelligence and how does it relate to the policy change?

Personal Intelligence is an opt-in feature that connects data from multiple Google apps-including Gmail, Drive, and Photos-to AI tools like Gemini and AI Mode search. The policy update aligns with its launch, acknowledging that personal content flowing through these AI tools can shape the ads users see.

How can users prevent their personal content from being used for ad targeting?

Users can disable Personal Intelligence in the Gemini app under Personal Intelligence controls, where connected apps like Workspace and Google Photos can be disconnected. The setting is also accessible via the Google account dashboard under Data & privacy, then Personalization settings, then Manage Personalized Recommendations in Search Services, and finally Connected Content Apps.

Does Google scan entire inboxes to build ad profiles?

No. A Google spokesperson told Ars Technica that the company does not scan entire inboxes to build ad profiles, does not use sensitive information to tailor ads, and does not share this data with third parties. Ads remain relevant to queries, response context, and inferred interests.

What should developers building on Google's AI stack assume about user content routed through Personal Intelligence?

Developers should assume any user content routed through Personal Intelligence might influence the ad ecosystem, even if Google filters sensitive categories. The update signals the boundary between user data and advertising inventory is more porous than earlier guarantees suggested.

Source: Ars Technica