AivexaNewsSearch
AI news for builders and product teamsChecked every hour

Updated Google privacy policy says your Gmail might be used for advertising

Collected Oct 23, 2026

Google updated its privacy policy on October 16, revising the section on ad personalization in a way that removes a blanket promise about sensitive app data. Android Authority spotted the change. The earlier text carried a plain commitment that personalized ads would not be based on content from Drive, Gmail, or Photos. The revised wording is more conditional, and in practice it allows content from those apps to feed advertising when that content has been shared with other Google services.

The mechanism that makes this more than theoretical is Personal Intelligence, which Google released in early 2026. Personal Intelligence lets a user connect data from several Google apps so that data can act as context for Google's AI products. It links to the Gemini app and also to AI Overviews and AI Mode in search. The premise is that richer context produces better answers. The consequence is that personal material flows into surfaces that also carry advertising.

Gemini itself does not currently display ads. AI-powered search does, and sponsored content can appear alongside AI search results. The policy revision appears to acknowledge that data routed into AI search through Personal Intelligence could therefore contribute to ad personalization. In other words, the protection around Gmail and Photos holds inside those apps but weakens once the data is shared onward.

Google disputes the harshest reading. In a statement, the company said connecting apps is entirely opt-in, that ads remain relevant to things like the query, the context of the response, and user interests, and that ads will not be based directly on a Gmail inbox or Photos library. It added that sensitive information will not be used to tailor ads. A Google spokesperson told Ars that the clarification does not signal a change in how data is handled with respect to advertising partners, that the company is not scanning entire inboxes to enrich ad profiles, and that none of this data goes to third parties.

The word doing the work in that statement is "directly." If a summary or insight drawn from your connected apps shapes an AI answer, and ads are tailored to the context of that answer, then your data has influenced ad targeting without being the direct basis for it. This suggests the practical effect is indirect influence rather than inbox-level profiling, which is a meaningful distinction but not a trivial one for privacy-sensitive users.

The broader pattern is one of tension inside Google's own design goals. Google wants its products to interoperate, and it treats the resulting pool of cross-service data as a strategic advantage for its AI efforts. Some apps, Gmail among them, retain protected status in Google's data handling. But each new connection is another route out of that protected space. The more a user shares across the ecosystem, the more exposure accumulates. Drive, Gmail, and Photos were long treated as a firewall zone for advertising; the update narrows that zone rather than eliminating it.

A related wrinkle concerns AI training. Google says it does not use personal data for training, but it does use user inputs and model outputs for that purpose. With Gemini wired into many services through Personal Intelligence, outputs can plausibly contain fragments of personal information. Google has said it tries to filter such content out of user activity controls so it stays out of training sets, but it does not promise the filtering always succeeds. In practice, that is an admission of residual risk rather than a guarantee.

Opt-in is real but soft. Personal Intelligence is enabled through prompts and pop-ups that encourage users to switch it on, and the consequences may not be fully spelled out at the moment of consent. Some users may have turned it on without realizing what it permits. For anyone who wants ads to stay clear of their personal data entirely, the lever is to keep Personal Intelligence off. The most direct path is the Personal Intelligence settings in Gemini, where connected apps can be disabled. The same controls are reachable through the main Google account settings: Data & privacy, then Personalization settings, then Manage Personalized Recommendations in Search Services, then Connected Content Apps. From there, connections to Workspace and Google Photos can be turned off.

For developers and product teams, the lesson generalizes beyond Google. When a platform lets users merge data across services to improve AI features, that data inherits the ad surfaces the AI touches. Consent flows need to make that chain visible, because a user who agreed to better answers may not have understood they were also agreeing to more relevant sponsored results. Teams building on Google's AI surfaces should treat connected-app status as a privacy variable they can inspect and explain, and should expect scrutiny of any feature that blends personal context with monetized output. The policy text now reflects that reality even if Google's public framing keeps the emphasis on what ads are not based on.

Read at Ars Technica · AI

Based on reporting from the original publisher. Visit the source for full context and later updates.

Publisher excerpt

Turning on Personal Intelligence gives the company more data for ad targeting.