Summary: A February 2026 report said Apple would open CarPlay to third-party AI chatbots such as ChatGPT, Claude, and Gemini in coming months, using a hands-free voice conversation mode. Siri would keep system tasks and the default slot, third-party apps would get no wake word, and vehicle and phone-control APIs would remain off-limits.
A Bloomberg report surfaced on February 6 through MacRumors, where Juli Clover laid out the news: Apple intends to admit third-party AI chatbot apps to CarPlay “within the coming months,” which would give ChatGPT, Claude, and Gemini their first official route onto the car’s interface. Historically the platform accepted only a narrow set of app categories, and no assistant other than Siri has ever held the voice slot. A day later, WinBuzzer followed with matching coverage.
The story here is not “a few more apps” — it is that the gate itself is opening. No computing environment polices attention as strictly as the car: eyes on the road, hands on the wheel. In that setting, a voice assistant stops being a nice-to-have and becomes the only practical interface. CarPlay has been a largely sealed surface since its 2014 debut; opening it gives Apple a fresh AI distribution lane outside the iPhone screen, fueled by commute time logged every day — and that directly moves usage frequency and retention for all three services.
The Assistant War Rolls Into the Car
The timing is no accident. This opening runs alongside another milestone on the iOS 26.4 software track: Siri is migrating onto a large language model, gaining the capacity to answer complex questions, execute multi-step tasks, and hold context together across apps. Apple is effectively playing both sides — upgrading its own assistant to face ChatGPT-class rivals head-on, while admitting those same rivals into the car, under conditions.
From Apple’s corner, this is textbook platform thinking: rather than let users fall back to dashboard phone mounts because Siri trailed behind, it pulls third-party AI inside an experience framework it owns, keeping control and safety review in house. For OpenAI, Anthropic, and Google, the draw is a high-frequency venue that used to be shut tight: American drivers on average spend over an hour a day in the car — minutes that, for now, go to radio, podcasts, and Siri.
Siri Keeps the Default Slot: Three Limits Apple Set
This is a partial opening, not a wholesale one. Read side by side, the MacRumors and WinBuzzer reports show Apple setting firm rules for third-party AI apps:
- Siri cannot be displaced. System-level jobs — texts, calls, music playback — remain Siri’s territory; a third-party AI enters as an add-on app, never as the in-car default.
- There is no wake word. The app has to be opened by hand before any interaction happens; nothing on par with a “Hey ChatGPT” callout exists.
- No control over the vehicle or the phone. CarPlay’s deeper control APIs remain off-limits to third-party AI apps — conversation is possible; acting on the car or the device is not.
Taken together, the pattern is plain: Apple shares the conversational value and keeps the system permissions for itself. A third-party AI can field questions, help shape a thought, or read a summary aloud — but launching navigation, adjusting the climate, and placing calls sit inside Siri’s moat.
Voice Is the Only Lane: Designing for Driving
The usage model the reports describe is deliberately spare: launching the ChatGPT or Claude app drops the session straight into a voice-conversation mode, with questions asked and answered hands-free. There is no text input and nothing to read at length. Those are safety-driven design trade-offs, and they define what qualifies for CarPlay: the AI experience that makes the cut reads as a phone call rather than a chat window.
For teams building AI products, this is constrained design at its purest. The screen displays almost nothing, voice is the single input channel, and the user is driving the whole time — so response length, pacing, and the rhythm of turn-taking all have to be worked out again from scratch. Move an experience honed for a desktop chat window onto a car interface, and it will almost certainly come up short.
Why Developers and Product Teams Should Care
Three consequences stand out. First, AI assistant distribution as a contest now formally reaches the cabin, giving ChatGPT, Claude, and Gemini a high-frequency, long-session, voice-only arena for standing apart from one another. Second, the way Apple has opened up — platform review, designated templates, no system permissions — is likely to serve as the reference point that other in-car systems, Android Automotive included, either emulate or push back against. Third, because the Siri rebuild and the third-party opening land together, the in-car bar for comparison climbs fast: the moment someone can put Siri and Claude side by side in one car, the gap is exposed on the spot.
This also fits a trend we sketched in our 2026 opening outlook, where the assistant race drifts from “which model is strongest” toward “which model is everywhere” — across operating systems, then browsers and earbuds, and now the automobile itself. For Apple, guarding the in-car entrance’s last meter plainly matters more than any leaderboard standing.