A privacy policy update effective October 16 rewrites the company's longstanding promise that it would never use content from Gmail, Drive, or Google Photos to personalize advertisements1,6. The new language adds a single, consequential exception: the pledge no longer holds if a user connects those services to Google's AI-powered tools through Personal Intelligence2,5. The change converts an opt-in AI feature into the legal mechanism by which Google's ad system could gain access to the most intimate data its users generate, collapsing the boundary between AI utility and ad targeting. The policy language is permissive, not directive—it authorizes a data flow rather than confirming one is live—but the authorization itself is the news: Google had never before carved an exception into this pledge, and the new text gives the company unilateral discretion over when and how to activate the pipeline.
The prior policy language was unambiguous. Since July 1, 2021, Google's privacy policy stated that it did not show personalized ads based on content from Drive, Gmail, or Photos. A line-by-line comparison of the October 16 text against the 54 earlier versions in Google's privacy policy archive shows that the new wording rewrites three sentences in the advertising section and adds a new explanatory note, with none of the earlier versions containing an equivalent carve-out. The updated text now reads that the restriction does not apply "unless you choose to make data from these services available to other Google services that support personalized ads".
The vehicle for that choice is Personal Intelligence, the opt-in feature Google released in early 2026 that allows users to hand data from multiple apps to Google AI products. The idea behind Personal Intelligence is that added context from a user's existing data leads to better AI answers. Google has been expanding its AI capabilities throughout 2026, including an AI-powered Gmail inbox that analyzes messages to highlight important information, reminders, and upcoming tasks. Each of those AI integrations now doubles as a potential data conduit to Google's advertising infrastructure. Whether the conduit is active today or reserved for future use, the policy language no longer prohibits it.
The structure of the change matters as much as its substance. Google frames the policy update as user-driven: the data flows only if a person "chooses" to connect services. But as 9to5Google noted, the opt-out mechanism is blunt. Users cannot selectively permit AI access to their email while blocking ad use of the same data; the only way to prevent ad targeting is to disable Personal Intelligence entirely7. Some AI features may function without Personal Intelligence enabled, which means users are not locked out of all AI utility. But Personal Intelligence is the integration layer Google is most aggressively promoting, and the features that depend on cross-service context require it.
ANALYSIS The result is a tiered trade-off rather than a pure binary, but the tier that delivers Google's most differentiated AI capabilities is the one that opens the ad door.
Google still maintains that advertisers do not get to see anyone's emails or photos directly. The ads people see can now be shaped by summaries and inferences drawn from that private content, but the raw data itself is not exposed to third parties.
ANALYSIS That distinction is real but narrow. The competitive value to Google lies not in selling raw email text to advertisers but in using it to build richer user profiles internally, profiles that make its ad auction more precise and its inventory more valuable.
The timing is notable. Google made the policy change on October 16 but did not announce it publicly; it was first spotted by Android Authority and subsequently reported by The Verge on October 233.
The competitive context sharpens the stakes. Google's ad business has always rested on the depth of its user data, but the company had drawn a visible line around certain categories of content. That line served a dual purpose: it reassured users and it preempted regulatory scrutiny.
ANALYSIS By tying the removal of that line to an AI feature that Google is actively promoting across its product surface, the company is betting that the utility of personalized AI answers will be compelling enough to override privacy hesitation, effectively using AI adoption as the consent mechanism for expanded ad targeting.
The strongest counterargument is that this is simply transparency. Google could have quietly used Personal Intelligence data for ad targeting without updating its policy; instead, it wrote the exception into the document. The change is opt-in, not default, and users who never enable Personal Intelligence are unaffected. That reading is fair as far as it goes, but it understates the dynamic. Google controls both the AI feature and the ad system. It sets the defaults, designs the onboarding flows, and determines whether the opt-in is presented as a prominent choice or buried in a settings menu.
ANALYSIS The formal structure of consent does not eliminate the asymmetry of design power. It is also possible that the carve-out exists primarily to enable AI features that require cross-service data access, with ad targeting as a legal byproduct rather than the driving purpose. But the policy text does not distinguish between the two uses, and Google has not publicly committed to keeping them separate. The absence of that commitment is itself the risk.
Ars Technica described Personal Intelligence as "proving to be a privacy minefield".
ANALYSIS The feature was introduced as an AI quality improvement; it is now also the legal gateway to a new class of ad data. Users who enabled it before October 16 made their choice under a privacy policy that did not contain this carve-out. Whether that gap is legally dispositive or merely optically damaging will depend on the jurisdiction.
What to watch: whether Google updates its Personal Intelligence onboarding flow to disclose the ad-targeting consequence explicitly, and on what timeline. Whether any data protection authority opens a formal inquiry into whether the retroactive application of the new policy to existing Personal Intelligence users meets prevailing consent standards. Whether Google's next quarterly earnings call references the expanded data access as a factor in ad revenue performance. And whether competitors that have built AI products around on-device processing use the policy change to sharpen privacy-focused differentiation. The wall between AI utility and ad monetization held for five years. Its removal will be tested by regulators, competitors, and users who may not yet know the terms have changed.