1. What Google changed
Google updated its privacy policy, effective October 16, to allow content from Gmail, Google Photos, and Google Drive to influence personalized ads when users connect those apps to other Google services, according to Search Engine Journal citing gHacks. For years, the policy stated that Google did not show personalized ads based on content from Drive, Gmail, or Photos. The revised language keeps that statement but adds an exception for cases where users choose to make data from these services available to other Google services that support personalized ads, as Search Engine Journal reported.
The change applies primarily to users who opt into Personal Intelligence in Gemini and AI Mode in Search. Google has begun testing ads within AI Mode for users who have Personal Intelligence enabled. According to Google's Search Help documentation cited by The Keyword, ads are not matched directly to raw email or photo content. Instead, Search may reference summaries, excerpts, and inferences from connected sources to personalize the experience, and the same information may be used to show ads, per Search Engine Journal.
2. How the signal pathway works
Personal Intelligence launched in January 2026 as a Labs experiment for Google AI Pro and AI Ultra subscribers in the U.S., according to Search Engine Journal. Users can connect Gmail and Google Photos through personal Google accounts. Workspace business, enterprise, and education accounts are excluded.
Google states that advertisers will not get access to a user's connected inbox or photo library. Sensitive categories (race, religion, sexual orientation, health) remain excluded from ad personalization. Users can disconnect apps or turn off personalized ads through My Ad Center. However, as Search Engine Journal noted, turning off ad personalization in My Ad Center does not necessarily stop connected apps from feeding data into AI features. Disconnecting the apps is the only way for a user to remove Gmail and Photos from Personal Intelligence entirely.
Google's help page also notes that connected-app summaries, excerpts, and inferences may be used to train Google's generative AI models, and a subset of data processed by AI Mode is reviewed by human reviewers, according to Search Engine Journal.
3. What advertisers can and cannot control
Search Engine Journal reported that there is no new targeting toggle, audience segment, or dedicated reporting inside Google Ads for Personal Intelligence signals. Connected-app inferences can influence which ads an opted-in user sees, but advertisers cannot directly target, segment, or bid on Gmail or Photos insights. Campaign eligibility follows existing AI Mode ad rules: Performance Max, Shopping, and Search campaigns using broad match (including AI Max for Search) are eligible.
Two reporting gaps exist, per Search Engine Journal. Google has not announced a report showing when connected-app information influenced an impression. And Search term, asset, and landing page reports for AI Max do not identify whether Gmail or Photos information contributed to delivery.
It remains unclear whether connected-app signals personalize ads only inside AI Mode or across other Google properties. Search Engine Journal observed that Google's help page answers the ad personalization question for Search Services as a whole, a group defined as Google Search (including AI Mode and Discover), Maps, Shopping, Flights, Hotels, Translate, and News.
4. Implications for enterprise privacy and data strategy
This development changes the calculus for enterprise teams in two directions. On the buy side, marketers running Performance Max or broad-match Search campaigns may see delivery influenced by private user signals they cannot observe, audit, or control. That opacity complicates attribution and makes it harder to explain ad performance to internal stakeholders. Our earlier analysis in always-on AI agents and data layer readiness anticipated this pattern: AI systems that consume private context create attribution gaps that existing reporting cannot close.
On the compliance side, the fact that Google infers intent from personal email and photo content, even with user opt-in, puts pressure on how enterprises think about privacy compliance within their own marketing stacks. If Google's ad ecosystem now treats personal communications as signal inputs, enterprise teams should expect regulators and customers to apply the same scrutiny to how first-party data moves through marketing automation platforms.
Consider auditing your current consent architecture and subscription center configurations to verify that your own data collection practices can withstand the heightened scrutiny this kind of industry shift invites. When a platform as large as Google blurs the line between personal utility and ad targeting, public sensitivity to data use rises across the board.
We recommend reviewing your Performance Max and broad-match campaign structures now for attribution clarity. Without dedicated reporting from Google, teams should establish baseline performance metrics before connected-app signals become widespread, so any shifts in conversion patterns or cost-per-acquisition can be isolated. This is a practical application of data management discipline: if you cannot see the input signal, you need stronger controls on your own measurement layer.
Consider pressure-testing your first-party cookies strategy and identity resolution practices. As we discussed in identity resolution across silos, the fragmentation of identity signals across platforms creates compounding risk. Google's move to infer ad relevance from private app data adds another layer of opacity between the signal and the served impression, making your own first-party data infrastructure more important as a source of measurable, governable truth.
The absence of advertiser controls is the most telling detail. Google is building a personalization engine that operates entirely on the consumer side of the transaction, invisible to the marketer paying for the impression. Enterprise teams should treat this not as a targeting opportunity but as a transparency deficit that demands stronger internal measurement and privacy assessment practices.
