Google Personal Intelligence May Use Shared Data for Ads

Connecting Gmail or Photos to Search could make that data a signal for targeted ads. Google Personal Intelligence supplies those connections, and Google's updated privacy policy now states the advertising trade-off in plain language. Here's how to check what's connected, pick an opt-out, and tell the default protection from the exception you opt into.

If you're a developer or IT pro using Google services for work, this decision goes beyond whether the AI answer is useful. The source is clear: data shared across services can feed ad personalization. Judging a connection only by the answers it gives you misses that second use entirely. The write-up on thehackernews.com covers the policy wording and user-facing opt-out options.

What changed in Google's privacy policy?

News of the policy update was first reported by Android Authority; the update lists an effective date of October 16, 2026. Google now spells out how letting Search or Gemini Apps pull data from your other Google services can personalize content, ads included, under that service's own rules.

The line concerns advertising: Gmail, Drive, and Photos content isn't used for personalized ads by default. Google's revised policy states:

“We don't show you personalized ads based on your 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.”

— Google's revised privacy policy, as quoted by The Hacker News

The default protection is still there. What changed is that the exception is now explicit: share content with another Google service and it can reshape how that content feeds ad personalization. Search is the named example: it may process whatever data you allow it to access in order to personalize content, ads included.

That's why the old line — "Google doesn't use Gmail content for personalized ads" — no longer tells the whole story. It skips the part where you choose to make that content available to a service that does serve personalized ads.

Follow the shared data, not just the app name

Personal Intelligence is the bridge between siloed services and personalized AI answers. Google rolled it into Gemini back in January 2026. The privacy trade-off the report flags is what happens once data from connected services lands in the service answering you.

The policy describes two different situations:

For Search, the connection is explicit: it may process some of what you allow it to see in order to personalize content, ads included. The report also notes that data feeding AI Mode or Gemini answers — with Personal Intelligence on — could turn into an ad signal.

Check the receiving service's permitted uses before connecting an app. The real question is what the destination service can do with that data once you hand it over.

Work accounts make this a permission decision

You use Gmail for project threads and Drive for team docs, and you're weighing connected features for better answers. That's a real benefit. But the policy forces a second question: will the data you share for those answers also feed personalized ads?

Wanting an answer drawn from your data is one thing; accepting what the receiving service is allowed to do with that data is another. They're different preferences, even when the UI bundles them into one toggle. A better answer doesn't settle the privacy question on its own.

My take: vet each proposed connection against what data it exposes, especially for a work mailbox. That's about scoping access narrowly where it makes sense — not because the source carves out a special workplace exemption. It doesn't.

This is exactly why we need transparency in AI: the explanation has to cover what happens to connected data, not just how tailored the answer feels. Google's policy hands you an advertising exception that belongs in that assessment.

Review Google Personal Intelligence connections and choose your opt-out

The report names two opt-out paths: disconnect individual apps from Personal Intelligence, or kill Personalized Recommendations entirely. Both are user-level settings. Start with the scope you're comfortable allowing, then pick the matching lever.

1. Open Google's Search Services Personalization settings. The source gives the address as www.google[.]com/search-personalization; replace the bracketed dot with a normal dot when entering it in your browser.

2. Review the apps you have connected to Personal Intelligence. Consider whether you still want their information available for personalised answers, given the policy's advertising exception. Do not base that decision solely on the default treatment of content inside Gmail, Drive or Photos.

3. Disconnect the individual apps you do not want to share through Personal Intelligence. This is the selective opt-out described in the report: choose the apps whose connections you want to remove rather than treating the feature as an all-or-nothing decision.

4. If you prefer the broader opt-out instead, disable Personalized Recommendations entirely through Search Services Personalization. Choose it when selective connections do not match the privacy boundary you want.

For a team auditing work accounts, a useful note would log the intended connection and the rationale for keeping or cutting it. That's a decision-making habit, not a Google config requirement. It keeps the conversation anchored to a concrete permission, not a vague stance on AI.

The report leaves ad mechanics and organisation-wide controls unspecified

The policy describes a possible use, not a guaranteed outcome for every interaction. The wording states that shared info may personalize content, ads included. That's reason to take the permission seriously. It's not proof that every linked email or photo spawns a specific ad.

• The report does not identify the exact fields extracted from connected content, explain how advertising signals are weighted, or provide a way to trace an individual ad back to a specific email or photo.

• The report describes user-facing opt-out settings. It does not document organisation-wide administration, account-management procedures, or a fleet-level control for enforcing the same choice across company accounts.

It also doesn't say disconnecting an app wipes previously processed data or rolls back earlier ad personalization. Treat the settings as your opt-out levers, not proof of a historical-data deletion guarantee.

The report doesn't show a new ad system flipping on when the policy took effect. What it gives you is an updated statement of permitted data use. It also doesn't establish that Gemini itself serves ads; just because Gemini appears in a connected-data flow doesn't mean it's showing ads.

If you don't enable these connections, the rule stays simple: the source says Gmail, Drive, and Photos content isn't used for personalized ads by default. The exception is you choosing to make that data available to another Google service that does support them.

Keep the answer benefit separate from the advertising permission

Personal Intelligence's pitch is tailored answers drawn from connected services. The revised policy makes clear that the same sharing decision can carry an ad consequence. The practical call is whether that combo fits your actual use case — not whether personalized answers are useful in theory.

For teams eyeing generative AI chatbots for security work, I'd settle the connected-data permission before linking a work mailbox. If the benefit is clear and the permitted use is acceptable, keeping a selective connection is a deliberate choice. If the trade-off isn't acceptable, cutting the connection is the consistent move.

Review the connections, then pick the opt-out that matches your boundary. Google's default protection and its sharing exception both live in the policy. A privacy decision has to account for both.