Google has begun testing Google AI Mode personalized ads for people who enable Personal Intelligence and connect supported apps to Search. Reported on October 16, 2026, the experiment brings ads into an experience that had been ad-free for these users.
The change could help Google select more relevant ads using personal context, including information derived from connected apps. It does not give advertisers access to Gmail inboxes, Google Photos libraries or a new audience they can target in Google Ads.
For marketers, the question is less about a new campaign setting and more about accountability. If extra context influences which ads appear, how can an advertiser evaluate the outcome when Google has not introduced dedicated controls or reporting for this test?
What changed in Google AI Mode personalized ads
Google’s official documentation on connecting apps to Search explains the boundary: Search services may reference summaries, excerpts and inferences from connected apps to personalize experiences, and that information may also be used to show ads. Google says ads will not be based directly on a person’s Gmail inbox or Google Photos library.
Previously, the connected-app Personal Intelligence experience in AI Mode did not show ads. Google is now testing advertising for some of these users, but has not announced a general rollout or a separate advertiser product.
| Area | Confirmed position | What advertisers should assume |
|---|---|---|
| User eligibility | People must choose to connect supported apps | The test does not cover every AI Mode user |
| Potential signals | Query, AI answer context, interests and derived connected-app information | Selection may account for context beyond the typed query |
| Data access | Advertisers cannot access connected Gmail or Photos content | No new customer-level data feed |
| Campaign controls | No new controls or targeting segments announced | Nothing new to activate in Google Ads |
| Reporting | No test-specific report announced | No way to isolate performance attributable to connected-app insights |
Google has not published the number of test participants, the exact campaign eligibility rules or a broader launch date. That limits what marketers can reasonably plan around AI Mode ads today.
Connected-app signals are not a new Gmail audience
The most important privacy distinction is between Google using derived information within its own service and an advertiser gaining access to that information. Those are not the same thing.
Google says users choose whether to connect apps and can disconnect them. It also says advertisers will not have access to Gmail inboxes or Google Photos, and that sensitive categories such as health or religion are not used for personalized ads.
That distinction does not mean connected-app information has no advertising use. A summary or inference drawn from connected content can still be relevant to personalization, even when the ad is not based directly on an inbox or photo library.
Google also describes using summaries, excerpts and inferences to improve Search services and train generative AI models. That is a separate data use and should not be confused with the information used to select an ad. Its help documentation does not explain the precise influence of connected-app context on individual ad impressions.
Advertisers still cannot build a segment of people who booked flights, opened an invoice or attended a meeting through connected apps. Personal Intelligence ads are a platform-side personalization test, not a permission to target someone’s personal files.
Ad delivery could change without adding advertiser controls
AI Mode advertising already has several paths into Google’s Search ecosystem. OneMetrik’s coverage of Search campaigns entering AI Mode through exact and phrase match describes a separate test that concerns campaign eligibility. Google AI Mode personalized ads are different: they concern the user context that may help select an ad.
Likewise, Google AI Max reporting updates involve search matching and transparency within advertiser-controlled campaigns. They do not provide a toggle for targeting Personal Intelligence users.
Advertisers cannot bid more specifically because someone connected Google Calendar or select an offer based on a related email. Google has not made those options available.
Nor does the test fix weak campaign inputs. If a SaaS advertiser optimizes for every form fill equally, adding personal context to ad selection will not tell the system which leads later become qualified opportunities. Cleaner conversion definitions and relevant offers remain more useful than guessing at undisclosed signals.
The B2B SaaS opportunity is narrower than the consumer story
Personal Intelligence lends itself naturally to consumer tasks. Travel bookings, recent purchases and preferences can make a recommendation more relevant in the moment. That is easier to picture than a complex enterprise software purchase, where budget approval, security reviews and several decision-makers shape the outcome.
Google’s connected-app features require a personal Google Account, and individual app connections vary by region. For example, the ability to connect Google Calendar to Search is currently limited to US English users. Our earlier coverage of Google Calendar context in AI Mode explains how schedule information can make an AI response more personal.
But personal context is not automatically company context. A user researching cybersecurity tools from a personal account might have relevant interests, yet that does not tell the advertiser their employer’s technology stack, procurement timetable or purchasing authority.
For B2B marketers, Google AI Mode personalized ads may help match some searches with more relevant offers. That is a possibility, not an observed improvement in SaaS lead quality. CRM audiences, account-level research and sales-stage feedback remain clearer inputs for reaching and evaluating business buyers.
Teams new to the surface can use our Google AI Mode guide for the distinction between conversational Search, AI Overviews and traditional results. In a separate Search update, Google is also making AI Mode source links more prominent, which affects content discovery rather than ad-targeting controls.
Better personalization creates a harder measurement question
Google has not announced a dedicated AI Mode ad reporting view showing when connected-app insights affected an impression, click or conversion. Advertisers therefore cannot separate results for users with Personal Intelligence enabled from other AI Mode traffic using a published test-specific metric.
A campaign CPA might improve while keywords, budgets, creative or conversion tracking also change. Without a credible comparison, that improvement cannot be attributed to connected-app personalization.
The distinction between organic and paid reporting also matters. Our analysis of Google’s AI Overviews reporting covers visibility data for AI-powered Search. That does not give advertisers a paid-media breakdown of this Personal Intelligence experiment.
Until Google offers clearer reporting, the sensible approach is to preserve a campaign baseline, log meaningful account changes and measure qualified opportunities rather than impressions alone. An unexplained improvement in lead volume is not evidence of better personalization.
Five practical steps before broader availability
There is no special setting to enable for Google AI Mode personalized ads. The useful preparation is work that improves campaigns regardless of whether the test expands.
- Audit conversion quality. Separate raw submissions from qualified leads, opportunities and customers. Send reliable downstream signals back to Google Ads where the integration supports them.
- Make landing pages specific. State the target audience, use case, pricing context, integrations and proof points clearly enough to support a high-intent decision.
- Review Search automation settings. Check negative keywords, brand controls, URL expansion and current AI Max configurations. Do not widen targeting solely to chase this experiment.
- Document privacy and data flows. Make sure marketing and RevOps can explain what first-party information they upload and which consent requirements apply. Google using its own connected-app context does not give the advertiser access to that data.
- Keep an experiment log. Record creative, bidding, landing-page and conversion changes so shifts in CPA or pipeline are not incorrectly credited to a test the account cannot isolate.
Our coverage of Google Ads Data Strength Uplift reinforces the measurement point: recovering more reported conversions is different from proving an increase in incremental business.
What would make this test matter more
Three developments would change the practical calculus. First, Google would need to clarify where the advertising test is running and which campaigns are eligible. Broad access to Personal Intelligence itself does not establish the geography or scope of the ad experiment.
Second, marketers would need reporting that separates relevant AI Mode activity from ordinary Search delivery or at least explains how to evaluate changes without overstating attribution. Third, Google could publish clearer user-facing explanations of how connected-app information affects a specific ad and how mistaken inferences can be corrected.
None is a confirmed rollout commitment. Until Google supplies clearer details, this test is worth monitoring, not restructuring a media plan around.
Frequently asked questions
Can advertisers target Gmail users through Google AI Mode?
No. Google has not introduced a Gmail-based audience or targeting option for advertisers. The test can use derived connected-app context on Google’s side, but advertisers do not receive access to Gmail inboxes or Google Photos libraries.
Are personalized AI Mode ads available to all advertisers?
Google is testing ads for some AI Mode users who opt into Personal Intelligence and connect supported apps. It has not announced general availability, a new campaign type or separate targeting controls for this experiment.
Can advertisers measure personalized AI Mode ad performance separately?
Google has not announced a dedicated report showing which impressions, clicks or conversions were influenced by connected-app context. Advertisers should avoid attributing campaign changes to the test without an identifiable measurement method.
OneMetrik Takeaway
Google has crossed a meaningful boundary by testing ads in a Personal Intelligence experience that was previously ad-free. The fact that connected-app information may inform ad selection deserves attention, especially because Google’s own documentation distinguishes direct inbox targeting from the use of summaries and inferences.
But the business case has not caught up with the product test. Advertisers have no new audience control, no dedicated reporting and no demonstrated lift in qualified pipeline from this feature.
At OneMetrik, we would leave budgets and campaign structures alone for now. Tighten conversion quality, maintain clear privacy practices and ask for better AI Mode measurement before claiming a win. More context in Google’s auction is only valuable to advertisers when it produces outcomes they can verify.