ProBackend
News & Analysis
5 days ago5 min read

Google’s AI search ads enter the connected-app experience

Google’s AI search ads enter the connected-app experience

Google is testing ads in AI Mode for some users who connect apps through Personal Intelligence. The change makes ad use of derived context more consequential—and leaves key details undisclosed.

The boundary has moved

An app connection that makes search feel more personal now sits closer to advertising, too. Google is testing ads in AI Mode for users who connect apps through Personal Intelligence, changing the practical meaning of an earlier ad-free assurance. The move is notable less because AI Mode has never carried ads—it has—than because the test reaches users whose connected sources can inform personalized answers.

Google’s published description says ads may be relevant to a query, the response context and a person’s interests. That does not tell us precisely which signal caused an ad to appear. Nor does it establish that every ad shown to a connected-app user draws on information from those apps. The test is real; the mechanics remain mostly opaque. (Search Engine Land: https://searchengineland.com/google-begins-testing-ads-in-ai-mode-for-users-with-personal-intelligence-enabled-496990)

What changed from the earlier position

Google previously said that people who chose to connect apps with AI Mode were not seeing ads at that time. Its wording was explicitly time-bound: it also indicated that ads could eventually work much as they do elsewhere in Search, remaining relevant to the query and response context. The current test appears to put that future direction into practice for at least some users.

That qualification matters. It would be inaccurate to say Google promised that connected-app users would never see ads. But there is still a meaningful expectation shift between “not now” and a test that brings advertising into the experience. For users, connecting personal sources can feel like a choice about answer quality; the new arrangement makes it more directly a choice about the context surrounding ads as well.

Google’s own explanation says Search can use summaries, excerpts and inferences from connected sources to personalize the experience, and that such information may also be used for ads. That describes potential use, not a claim that every response or ad is built from app-derived information. (Google: https://blog.google/products-and-platforms/products/search/personal-intelligence-ai-mode-search/)

Raw content and derived context

The distinction Google draws is between direct use of raw Gmail or Photos content and use of information derived from connected sources. The company says ads will not be based directly on a user’s Gmail inbox or Photos library, while also saying summaries, excerpts and inferences may inform personalization and advertising.

Those statements can both be true, but the distinction is not self-explanatory. A summary or inference can carry meaning from a message or image without exposing the original item. For a user, that may still feel like personal information shaping an ad. Google’s public wording sets a boundary around direct use; it does not give a technical account of what qualifies as an inference or how such derived context is applied to a particular impression. (The Verge: https://www.theverge.com/tech/1017600/google-personal-intelligence-privacy-policy-ads-update)

Google also says advertisers do not get access to connected apps. That is an important separation: the company’s stated model is not that an advertiser opens a user’s inbox or photo library. But this assurance does not answer every question about Google’s own processing, how long derived signals persist, or what campaign reporting reveals. Those details are not established in the available descriptions.

Controls are useful, but incomplete

Google says users decide whether to connect content apps to Search Services and can change settings or disconnect apps at any time. That gives people a direct control over whether those sources remain connected. Users who do not want connected-source information shaping Search personalization can review those settings and disconnect the apps.

There is a practical gap, though: the published explanation does not specify whether disconnection immediately clears all previously derived context from systems involved in ad delivery. The existence of a disconnect control should not be mistaken for a documented deletion timeline.

Google also says it does not show personalized ads based on sensitive categories including race, religion, sexual orientation or health. That is a stated safeguard, but it should be read narrowly. The cited language names categories and a rule for personalized ads; it does not explain how every possible inference or edge case is handled.

What advertisers can—and cannot—infer

For marketers, the tempting conclusion is that connecting Gmail or Photos creates a new targeting opportunity. The available information does not support that. Google says advertisers cannot access connected apps, and it has not publicly detailed advertiser controls, whether campaigns can identify this setting, or whether the test changes reporting.

The more defensible reading is that Google is exploring how ads fit into an AI answer experience when the user has opted into personal context. Ads may be assessed against the query and response context, but advertisers should not assume they can target private app contents or buy a distinct Personal Intelligence audience. There is no disclosed campaign-level evidence here to support that claim.

For publishers, the test is another sign that conversational search can bring monetization into a response-oriented interface rather than leaving ads confined to familiar link lists. What that means for referrals, user trust or publisher economics is still unknown. The reports cited here supply no performance results or traffic measurements.

The questions the test leaves open

“Testing” tells us little about scale. The available reports do not establish the number of users, geographic reach, duration, eligibility rules, formats, or whether Google plans to expand the test. They also do not explain how an app-derived summary relates to a specific ad impression, or whether users can tell when connected context played a role.

The next useful disclosure would be concrete: what users see, what controls apply, what advertisers can measure, and what happens to derived information after disconnection. Until then, the narrow conclusion is the sound one. Google is testing AI Mode ads for some users who connect apps through Personal Intelligence; it says raw Gmail and Photos content is not used directly for ads, advertisers do not access connected apps, and users can disconnect them. The precise boundary between personal context and ad relevance remains unclear.

Sources

the boundary has movedthe boundary has moved