Gemini on Android Auto now reads your Gmail, Photos, and Calendar
Google is rolling out Gemini Personal Intelligence to Android Auto starting today, giving the in-car assistant opt-in access to Gmail, Photos, Calendar, Maps, and YouTube history.

Updated
Why it matters
- Google is rolling out Gemini Personal Intelligence on Android Auto starting today, granting the assistant access to Gmail, Calendar, Photos, and other services.
- The feature is opt-in and off by default, with per-app toggles for Google Photos, Maps, Search, YouTube, and Google Workspace.
- Google says Personal Intelligence data stays in its secure infrastructure, is not shared with third parties, and is not used to train foundational AI models.
- Google first launched Gemini on Android Auto and cars with Google built-in late last year; Elyse Betters Picaro tested Personal Intelligence on phones in April.
Starting today, Google is rolling out Gemini Personal Intelligence on Android Auto, giving the AI assistant access to your Gmail, Calendar, Photos, and other native Google services while you drive. The feature is opt-in, off by default, and lets drivers choose exactly which apps Gemini may tap.
The rollout marks the second phase of Google's push to embed Gemini in vehicles. The company first brought the assistant to Android Auto and cars with Google built-in late last year, replacing the older Google Assistant. Now the assistant gets what it was missing on the road: your personal context.
The stakes here are straightforward. Voice assistants in cars live or die on how few words it takes to get a useful answer, and Google is betting that an AI which knows your flight details, your dietary preferences, and your commute history will need fewer clarifying questions than one running on general knowledge alone. That is also a bet that drivers will accept an AI reading their email from the driver's seat.
What does Personal Intelligence actually do?
Google first launched Personal Intelligence on Android phones and in Search. The system lets Gemini retrieve and synthesize information across your personal apps to tailor its answers. ZDNET's Elyse Betters Picaro tested it in April and found the tailored results very helpful.
The Android Auto version works on the same principle, just at 65 miles per hour. With access granted, Gemini can pull your upcoming schedule, flight details, or stray photos — think receipts and concert posters — to sharpen its responses.
Google's own examples show the pitch. You can ask Gemini which gate your upcoming flight departs from and whether there are delays. Ask for restaurant recommendations nearby, and Gemini can check your search history and emails, notice you're vegan, and filter its suggestions accordingly.
If Personal Intelligence stays off, Gemini's responses will continue drawing on general knowledge only. Nothing changes for drivers who ignore the feature.
Can you trust it with your inbox?
Google has stated previously that all information processed through Personal Intelligence stays within the company's existing secure infrastructure and is not distributed to third-party services. Gemini also does not train its foundational AI models on personal context, according to the company, so your data is reserved for your own use.
The controls are granular. You can opt in and out at any time, and if you've already granted access on your Gemini profile, that permission carries over to the car. Drivers can toggle access for specific apps and groups, including:
- Google Photos
- Google Maps
- Search and Discovery Data
- YouTube and YouTube History
- Google Workspace (Gmail, Calendar, and Drive)
You can also delete stored memory items, such as past conversations with Gemini. That matters if you'd rather the assistant not resurrect the time you asked about the blinking symbol on your dashboard.
Why the car changes the equation
The core argument for Personal Intelligence on Android Auto is focus. Swapping Google Assistant for Gemini already made the driving experience less frictionless in one respect: users can talk to the software in natural language instead of memorizing command phrases. More context should push that further, meaning fewer follow-up questions, less clarification — because, as ZDNET's coverage notes, Gemini is not always perfect — and more attention on the road.
The concrete use cases are the kind of thing drivers have wanted for years. ZDNET's author points to music results based on the artists you actually listen to, or finding a gas station near Newark Liberty instead of JFK because you've only flown out of one airport in the past year. Those are small conveniences. They are also the difference between an assistant that feels like it knows you and one that feels like a search box.
There are open questions. Upgrading the AI's context should fix most of those problems, ZDNET's author writes, though the piece adds: "I'm not holding my breath for a near-flawless exchange every time." The author also flags whether digging deeper for information will noticeably delay Gemini's responses — a real concern on the road, where cellular signal is not always reliable.
The bigger picture
This rollout fits a pattern. Google has spent the past year moving Gemini from phones into Search, Workspace, and now the car, with Personal Intelligence as the connective tissue. Each surface increases both the utility and the amount of personal data the assistant touches, which makes the opt-in design and per-app toggles the load-bearing part of the privacy story.
For now, the feature's success will come down to whether tailored answers are fast and accurate enough at highway speeds to justify handing Gemini the keys to your inbox. Drivers who want to test that can turn it on today, one app at a time.
Original: blog.google
More from Elena Vasquez
Show full bio
Market editor covering technology and product news in artificial intelligence and machine learning at AI In Context.
305 articles