Google brings Gemini to cars
- Google started rolling Gemini into cars with Google built-in on April 30, replacing Assistant in new and existing vehicles via software updates. - The rollout starts with U.S. drivers using English, and Gemini can answer owner’s-manual questions, handle follow-ups, and use Gemini Live hands-free. - This matters because Google is moving from phone projection to built-in car software — and latency now becomes a product problem.
Cars are turning into software platforms, not just machines with screens. That changes what a voice assistant is supposed to do. It’s no longer just “call mom” or “play this song.” It has to understand the route, the cabin, the car’s own settings, and sometimes the owner’s manual too. That’s the backdrop for Google’s latest move — on April 30, it began rolling Gemini into cars with Google built-in, replacing Google Assistant in both new and existing vehicles through over-the-air updates, starting in U.S. English. (blog.google) ### What’s actually changing in the car? The big shift is that Gemini becomes the built-in assistant inside vehicles running Google’s in-car software, not just a phone app mirrored onto the dashboard. Google says drivers can speak more naturally, ask follow-up questions, control vehicle settings, manage messages, find stops, a(blog.google)igger deal than it sounds — it means the installed base matters, not just new model launches. (blog.google) ### Wait — is this Android Auto? Not exactly. Android Auto runs on your phone and projects onto the car display. Android Automotive — often marketed as Google built-in — runs directly on the vehicle hardware. That distinction matters because a built-in assistant can go deeper into the car itself. It can tie into infotainment, (blog.google)but this rollout is about the built-in one. (source.android.com) ### Why does GM keep coming up here? Because GM has been one of the clearest examples of this strategy. The company already bet on Android Automotive for newer infotainment systems, and outside reporting says Gemini is headed into GM vehicles in the coming months under the same Google built-in framework. So this is not a random feature drop — it fits a longer(source.android.com)ar stack. (gmauthority.com) ### What can Gemini do that Assistant really couldn’t? Mostly, it reduces command friction. Old car assistants worked best when you spoke in the exact shape they expected. Gemini is supposed to handle messy human speech — “find a lunch place on the way, not fast food, with outdoor seating” — and keep the conversation going. Googl(gmauthority.com)ore of a conversational layer. That sounds subtle, but in a car, fewer retries matters a lot. (blog.google) ### So where’s the catch? Latency. Cars are one of the worst places for an AI assistant to feel slow or confused. A recent Google Home update tried to speed up Gemini’s smart-home command handling by as much as 1.5 seconds, and even that small number tells you something important — real-world environments with lots of devices, (blog.google)t the environment is harsher: noise, motion, interruptions, weak connectivity, and higher stakes. (cepro.com) ### Why is Google pushing this now? Because the platform underneath is getting broader. In March, Google said Android Automotive is expanding beyond infotainment toward a more centralized software-defined-vehicle foundation for non-safety functions. Basically, the dashboard assistant is becoming one piece of a larger car(cepro.com)r for the vehicle itself. (blog.google) ### Bottom line? Google isn’t just putting a chatbot on a dashboard. It’s testing whether conversational AI can become the default interface for software-defined cars. The promise is obvious — less menu-diving, more natural control. But the real test is boring and brutal: speed, reliability, and whether the system still works cleanly when the road gets noisy. (blog.google)