General Motors did not merely remove two familiar icons when it began dropping Apple CarPlay and Android Auto from new EV programs with the 2024 Chevrolet Blazer EV. It took greater responsibility for the useful life of the dashboard. Applications that could follow an owner from one replaceable phone to the next would instead depend on the car’s embedded system. Integration might improve. Choice certainly narrowed. GM can make an attractive interface for a new car; the risk appears later, when continued usefulness depends on GM maintaining the embedded system.
Losing a connected service does not disable the whole vehicle, but it can permanently remove functions from intact hardware. Toyota’s retirement of 3G Safety Connect illustrates the smaller, real injury. From November 2022, affected vehicles, including suitably equipped 2010–2019 4Runners, lost services such as automatic collision notification and the emergency assistance button. Navigation remained available. The cars did not stop driving. Intact hardware had nevertheless lost a function for which somebody had paid, because a communications network outside the owner’s control had reached the end of its life.
Phone projection reduces one part of that exposure. CarPlay and Android Auto let a phone supply applications and much of their update cycle, although they still need a compatible, functioning head unit and connection. Android Automotive OS is different: it runs in the vehicle. Google built-in services and a cellular data plan are different again. An embedded system can make good use of battery information, charging stops, and other vehicle data. That is a reason to develop one, not an explanation of why a customer must lose the alternative. Excellent integration and continued choice need not be enemies.
Nor is every electronic failure a death sentence. A dead display, a discontinued subscription, a failed controller, and a closed authentication server are different problems. Some remove conveniences; some make essential functions inaccessible. The repair question is whether a replacement exists, whether it can be paired, and whether the required software and diagnostics remain available. An alternator is a useful comparison because it can usually be replaced, not because older cars were collections of perfectly independent parts. The important distinction is between an expensive repair and a repair no one is permitted or equipped to perform.
Porsche has supplied a concrete answer for a limited part of that problem. Its PCCM and PCCM Plus units put modern communications into older dashboards without asking owners to discard the cars around them. The 2023 expansion covered, among other models, specified 997-generation 911s and 987 Boxsters and Caymans. This is not a promise to reproduce every control module forever. It is something more useful than a promise: a replacement product with defined compatibility, designed for the space and peripherals already in the car.
A classics department cannot keep electronics useful by itself; a certificate of authenticity will not repair a navigation computer. Continued use depends on continued support: parts that fit, software that recognizes them, documented procedures, and someone willing to supply the service. Those provisions determine whether the car remains fully usable after the launch presentation is forgotten. A manufacturer that asks customers to trust its ecosystem also assumes responsibility for explaining how owners can exit it when services end.
Tesla, often cast as the leading exponent of the disposable electronic car, also offers a counterexample. Eligible older Model S and Model X vehicles can receive a paid infotainment hardware upgrade. The program has eligibility limits and equipment consequences, so it is not a cure for every future support problem. It does establish that replacing the computer rather than the car is a real commercial option. Durable design and ongoing electronics development are not mutually exclusive. They become incompatible when the manufacturer declines to connect them.
GM’s own position now requires more care than the original rejection headline suggests. On September 15, 2026, it announced a new interface for the 2027 Chevrolet Silverado and GMC Sierra light-duty pickups that accommodates CarPlay and Android Auto alongside native vehicle functions. Those are announced future pickups, not a retrofit for every existing GM EV. Still, the proposed coexistence is revealing. Phone projection can occupy the dashboard without preventing GM from displaying trailering information or developing an interface of its own. The choice is architectural, not theological.
A credible support commitment would therefore describe which important functions remain available offline, how long services and security support are promised, and what happens when a component or network disappears. It would address replacement hardware and pairing, not simply celebrate over-the-air updates. Updates are valuable while they arrive. An owner keeping a car for 20 years needs to know what the plan is after the last one.
There is a commercial incentive to keep customers inside a manufacturer’s interface, including the opportunity to sell continuing services. That does not prove an intention to disable old cars. GM gains immediate control over the interface, while the owner carries the later risk that services, networks, authentication, or replacement hardware disappear. Removing an independent route to maps, music, and communication makes a clear support policy more important, not less.
A satisfactory old car cannot depend on a permanent relationship with its original salesman. GM can choose its software strategy, but it cannot fairly present control as a customer benefit while leaving the customer to absorb an undefined service life. The failure occurs when a still-functional car loses ordinary maps, music, communication, or safety services because the embedded ecosystem ages out. The dashboard becomes the place where the owner keeps finding receipts for things the car can no longer do.
