Google says Gmail data is not used for ads now, but AI has reopened the question
Google is not currently using Gmail data for targeted advertising, according to statements cited in a new report. But the company’s expanding AI features are raising a more consequential question: whether the line Google drew years ago around email-based ad targeting can hold as its products become better at synthesizing personal information across services.
The issue matters because the underlying technology is changing faster than the policy language that surrounds it. Since 2017, Google has said it would not show ads based on users’ Gmail messages. That promise helped establish a boundary around one of the most sensitive forms of personal data the company handles. The new concern is not that Google has formally reversed that commitment. It is that AI systems can now generate broader user insights from inboxes and other personal sources in ways that make the old boundary less clear in practice.
As described in the report, Google’s support documents already acknowledge that ads may be shown based on users’ interactions with AI. At the same time, the company has introduced and expanded so-called Personal Intelligence features that can draw on data from Gmail, Google Calendar, and Google Photos. Put together, those two facts create a new privacy and advertising dilemma even if no policy has changed today.
A Google spokesperson said by email that the company is not using Gmail data for targeted ads right now. The spokesperson also pointed to a statement indicating that users who enable Personal Intelligence features currently will not see ads in the AI Mode of Google Search. When asked whether Personal Intelligence is being used to target ads elsewhere, including regular Google searches, the spokesperson said that is not the case today.
The repeated qualifier is the key phrase: today. The report argues that neither the company response nor the public statement rules out future ad targeting based on insights derived from personal data. That does not prove a policy shift is coming, but it does show why observers are scrutinizing the wording so closely. In the AI era, the distinction between raw data and inferred knowledge can become harder to defend.
AI changes what can be inferred from ordinary digital traces
The concern is not simply that Google can read information from a user’s inbox. The bigger shift is that AI can combine details across services and turn them into a much richer profile of intent, habits, and likely preferences.
The examples in the report are straightforward and revealing. A user researching new cars could receive recommendations shaped by what Google infers from email, such as what vehicle they currently drive, what activities they might need a car for, and who else they may be driving around. Someone shopping for clothes could be analyzed through purchase receipts in Gmail to estimate preferred brands and styles. A traveler asking for destination ideas might prompt the system to review prior emails and photos to infer where they have already been.
None of those examples necessarily mean ads are being served from Gmail content today. But they show how AI compresses a chain of interpretation that used to be much weaker. Instead of matching an ad to a keyword in an email, an AI assistant can build a synthesized portrait from many signals and then use that portrait to shape responses, recommendations, or potentially ad relevance elsewhere in a company’s ecosystem.
That is what makes the current moment more complicated than older debates about email scanning. The practical value of the data is no longer limited to isolated facts. AI allows platforms to derive context from combinations of messages, calendar entries, images, and search behavior. The resulting insight can be more commercially useful than any single source on its own.
Why the wording around “currently” matters
Google introduced Personal Intelligence in January and made it available to all U.S. users in March, according to the report. Once activated, it allows the company to reference personal information from Gmail, Google Calendar, and Google Photos when users search in the Gemini app or in Google Search’s AI Mode. That rollout is significant because it moves personal-data-assisted AI from a limited experiment toward a mainstream consumer feature.
At the same time, the company’s public posture remains cautious. The report notes that ads are expected to come to AI Mode even for users who have set up Personal Intelligence. That does not establish that Gmail-derived insights will be part of ad targeting. It does, however, show that monetization and deeply personalized AI are now moving closer together in product design.
This is why privacy concerns are likely to intensify even before any formal change occurs. Once a company can use personal data to improve recommendations and conversational answers, users and regulators will naturally ask where that same intelligence can and cannot travel. Can it stay confined to assistance features? Can it inform commerce features? Can it shape ad relevance in adjacent products? The technical capability tends to outrun the policy categories people are used to hearing.
For Google, there is also a trust problem baked into the issue. Gmail has long been treated by users as infrastructure for intimate, financial, and logistical parts of daily life. Even if a company never places a display ad directly “based on your email” in the old sense, consumers may not see much difference if AI-generated profiles built from inbox content start affecting what products they are shown somewhere else.
That is why the present dispute is less about proving a hidden ad system exists and more about defining a credible boundary before one does. Google’s position, based on the available reporting, is clear on the present tense: Gmail data is not being used for targeted ads right now. The harder question is whether AI-driven personalization makes that promise increasingly difficult to interpret, enforce, or preserve over time.
For now, the company is asking users to separate what its AI can know from what its advertising systems can use. As AI products become more integrated with email, calendars, photos, and search, that separation may become one of the most consequential trust tests in consumer technology.
This article is based on reporting by Fast Company. Read the original article.
Originally published on fastcompany.com
