General Motors is betting that artificial intelligence can make time spent in the car more useful, and eventually let the vehicle handle chores for its owner. The automaker laid out the plan on Wednesday at its annual GM Forward event in New York City.
The company noted that Americans spent about 95 billion hours behind the wheel last year, covering roughly 2.76 trillion miles. GM does not aim to shrink that figure, Chief Product Officer Sterling Anderson told WIRED. Instead, he said the goal is to "make it more productive, make it more relaxing, make it more engaging," and to make the vehicle "a more central, more useful part of your life than it has been historically."
The first concrete step is OnStar AI, a new in-car assistant that will debut in 2027 Cadillac Lyriqs. The name comes from GM's 30-year-old roadside assistance and security service, which has 12 million subscribers. Working with a companion app, the system will initially handle simple requests, such as turning on the air conditioning by voice without digging through menus. Anderson said it could later coordinate school pickups and drop-offs based on which spouse's car is closer.
GM also intends to use AI to speed up how it designs and builds new vehicles. It aims to eventually cut its current development timeline of four to five years in half, which matches the pace at which Chinese automakers bring new models to market today.
On driving technology, the company plans to bring its "eyes-off" highway system to its most popular vehicles. GM previously said the feature would arrive on the Cadillac Escalade IQ in 2028. Its existing Super Cruise system already lets drivers take their hands off the wheel on hundreds of thousands of miles of pre-mapped highway. GM is not yet publicly discussing full autonomy. "This industry is rife with promises that were not fulfilled. I am not keen on making them," Anderson said.
He contrasted GM's approach with Tesla's "Full Self-Driving (Supervised)" feature, which can handle many highways and local roads but requires drivers to remain alert and ready to take over. Anderson, who once led Tesla's Autopilot program, said such a system "presents the illusion of autonomy without its benefits." If drivers cannot do other things while the car does most of the work, he argued, the system is less useful and risks the safety of occupants and others on the road.
GM instead plans to widen eyes-off driving gradually: first on certain highways, where drivers could text, eat or watch YouTube, then on more highways, then on local roads, and possibly everywhere. In a blog post, the company said the cabin could become a space to "relax, work, and unwind," paving the way for cars that handle everyday errands such as picking up groceries, dropping off children and perhaps finding a parking spot.
Urban planners and infrastructure experts have raised concerns about that vision. If cheaper, less burdensome trips lead people to send their cars out for every task, roads could fill with vehicles carrying sleeping or distracted passengers, or with no one aboard, and cities could sprawl further.
Anderson agreed with the premise. With cheap personal driving robots, he said, "you will very likely incur more of those robots on the roads," raising societal and infrastructure considerations. He argued there is time to address them, since cities sprawl slowly. "I wouldn't call this an imminent thing, but it certainly is something we should be thinking about," he said.
