1. What changed in the Google privacy policy update?
Google revised its privacy policy with an effective date of October 16, 2026, adding a conditional exception to a promise it has long made: that it does not use the contents of Gmail, Google Drive and Google Photos to personalize the ads people see. Under the new wording, material from those three services can influence advertising when a user connects it to Google's AI features. The edit is small on the page, but it touches one of the company's most recognizable privacy commitments.
The older language was simple. Content stored in the three products was treated as off limits for ad targeting, and that line was one of the clearest reassurances Google offered to people who keep personal correspondence, documents and family pictures on its servers. The updated text keeps the headline statement but attaches a qualifier. Google says it does not show personalized ads based on that content, unless the user chooses to make the data available to other Google services that support personalized ads.
That qualifier is the whole story. A flat rule has become a rule with a condition, and the condition depends on a choice the user makes elsewhere in their account settings. Nothing in the revision announces itself as a reversal, which is why the change drew attention only after the policy text was compared against its earlier version.
2. How does Personal Intelligence connect to ads?
The exception is tied to Personal Intelligence, the name Google gives to its effort to make its AI tools more useful by letting them draw on a person's own information. With the feature, AI Mode in Google Search and the Gemini assistant can reach into connected apps, including Gmail, Drive and Photos, to answer questions with personal context rather than generic knowledge.
The practical appeal is easy to see. Someone who asks an assistant about an upcoming trip can get an answer that accounts for the hotel reservation sitting in an old email, a document with an itinerary, or a photo of a confirmation screen. Without access to those sources, the assistant can only offer general suggestions. With access, it can behave more like a personal aide that already knows the details.
The revised policy shows the other side of that arrangement. Once the data has been connected so that AI features can use it, the same connection can feed the systems that decide which ads are shown. In other words, the permission that unlocks a tailored answer may also unlock tailored advertising, unless the user takes a further step to prevent it.
The arrangement reflects a wider tension in consumer technology. Assistants become more helpful as they learn more about the person using them, yet every additional data source raises questions about where that information travels and what other purposes it may serve. Companies typically resolve the tension through consent screens and settings, and the quality of those controls often matters as much as the policy text itself.
3. What limits does Google say it applies?
Google has described several guardrails around the exception. Using content from Gmail, Drive and Photos for ad targeting is off by default, so a user who has connected nothing and changed nothing is not affected. Advertisers do not receive access to anyone's emails, files or photos, which means the underlying content is not handed to the businesses that buy ad space.
The company also says it will not share personally identifying details with advertisers without permission. That distinction matters in the advertising industry, where personalization usually relies on signals and inferred interests held inside the platform rather than on the raw material itself. A system can learn that a person is planning a trip and show relevant offers without ever disclosing the booking confirmation to a third party.
These assurances narrow the scope of the change, but they do not remove it. The policy language still allows connected content to inform ad personalization, and the safeguards describe what Google will and will not pass along rather than whether the data participates in targeting at all.
4. Is Google reading private email and photos to sell ads?
Not wholesale, based on what the policy and the company's statements describe. The default remains that content from the three services is not used for ad targeting. The exception applies to people who connect their data to the AI features and who have not blocked its availability to other ad-supporting Google services.
Even so, the trade-off is real for users who embrace AI conveniences. A person who links an inbox and photo library to get better assistant answers may, by doing so, accept a loss of privacy in how ads are chosen. The cost of a more capable assistant is paid in data access, and the revised policy makes clear that advertising is among the places where that access can have an effect.
The distinction between default behavior and opt-in behavior is the central point for readers weighing the change. Because the targeting is off by default, the burden falls on people who actively connect their accounts to understand what that connection implies. Many users click through permission prompts quickly, and a policy exception that depends on such a prompt may not be noticed at the moment it applies.
Clarity is a recurring concern in such cases. A privacy policy is a legal document, and its conditional clauses can be hard to parse for people who are not accustomed to reading them. A statement that ads are not personalized from certain content, followed by an exception that depends on how other services are configured, asks a lot of a casual reader. Plain-language explanations at the moment of connection would help users judge what they are agreeing to.
5. How can users opt out?
Users can disconnect the apps from the AI features in their Google Account. The control sits under Data and privacy, in a section labeled Connected Content apps. Removing the connection cuts the link between the stored content and the services that would otherwise draw on it. Users can also switch off personalized recommendations, which limits how ad personalization works more broadly.
There is one caveat worth knowing. Information collected before a user disconnects may remain visible in their Search history. Disconnecting stops future use of the connected apps by these features, but it does not necessarily erase everything the services have already gathered. People who want a cleaner break may also need to review and delete their activity history separately.
For anyone who has not used Personal Intelligence at all, the practical advice is simpler. Check the Connected Content apps page to confirm that no accounts are linked that the user did not intend to link, and review the personalized recommendations setting to make sure it matches their preferences.
It is also worth remembering that connected apps and ad settings are separate switches. A user might disconnect Gmail from the AI features yet leave Drive or Photos linked, or might keep every connection and still turn off personalized recommendations. Reviewing each control individually is the only reliable way to be sure the combination reflects what the person wants.
6. Why does the timing matter for Google's AI push?
The revision arrives as Google works to place AI at the center of its products, from Search to the Gemini assistant to the apps people use every day. Personal data is the raw material that makes such assistants distinctive. Generic models are available from many companies, but a model that can read a user's own mail, files and pictures offers something competitors without that data cannot easily match.
That gives Google a strong reason to encourage people to connect their accounts, and it also explains why advertising is a sensitive subject. Google's business has historically depended on ads, and personalization improves the value of each impression. A policy that lets connected content contribute to that personalization aligns the AI strategy with the company's primary source of revenue.
The quiet rollout also fits a pattern of recent privacy disputes involving Google's AI efforts. In May, the Chrome browser was found to have installed a multi-gigabyte on-device AI model on users' computers without prominent notice, and the browser also dropped a privacy claim that had been attached to its Gemini Nano model. Both episodes drew scrutiny, and the latest policy edit adds to a list of changes that arrived with little fanfare.
Advertising personalization works by matching signals about a person's interests and circumstances to the ads most likely to be relevant. Content from an inbox or a photo library is rich in such signals, covering travel plans, purchases, hobbies and family events. That richness is exactly why the policy language matters, and why the guardrails around default settings and advertiser access carry so much weight in judging the change.
7. What should users watch next?
The policy took effect on October 16, so the exception is already in force. The next developments to follow are whether Google clarifies how connected content is weighed in ad personalization, whether it changes the prompts that appear when users connect their apps, and whether regulators or privacy advocates press for stronger disclosure of the new wording.
For now, the safest course for privacy-conscious users is to treat the connection of Gmail, Drive or Photos to any AI feature as a decision with advertising consequences, not only a convenience. The settings to review are the Connected Content apps list and the personalized recommendations switch, both found in the Data and privacy area of a Google Account. Those controls determine whether the exception applies to a given person, and they remain in the hands of the user.