General Motors is preparing to move one of its most ambitious technologies out of the luxury showcase and into the vehicles that form the heart of its American business. GM says its planned autonomous highway-driving system, designed to let occupants take their eyes off the road under supported conditions, will eventually spread from the Cadillac Escalade IQ to high-volume full-size SUVs and light-duty pickups.
The Escalade IQ remains the launch vehicle, with GM targeting 2028 for its first deployment. What changed is the scale of the plan. Instead of treating “eyes-off” driving as an expensive Cadillac experiment, GM now sees it as technology that can reach some of its best-selling Chevrolet and GMC vehicles. The company has not, however, given those additional models a specific rollout date.
Cadillac Is the Starting Point, Not the Destination
When GM first detailed its eyes-off strategy in 2025, the Cadillac Escalade IQ made sense as a technological launch pad. It is a high-end electric SUV with enough price and prestige to absorb costly new sensors, computing hardware and software before those technologies spread farther down the lineup. GM said the system would debut in 2028, initially focusing on highway driving where traffic flows are more predictable than on crowded urban streets.
The latest announcement makes clear that Cadillac is only the first stage. GM now says autonomous highway driving will expand into its highest-volume vehicle lineups, specifically naming full-size SUVs and light-duty pickups. It has not published a model-by-model schedule, so that does not mean a Silverado, Sierra, Tahoe, Suburban or Yukon is guaranteed to receive eyes-off capability in 2028. Instead, GM is describing a platform strategy: prove the technology first, then use a common computing architecture to bring it to far more vehicles without engineering an entirely separate autonomous-driving system for every model.
“Eyes-Off” Changes the Driver’s Job
Today’s Super Cruise can already make a long highway trip feel very different from conventional driving. On compatible roads, the system can handle steering, acceleration and braking while allowing the driver’s hands to remain off the wheel. But there is an important condition: the person behind the wheel still has to watch the road. A driver-facing monitoring system exists specifically to make sure that attention does not drift away for too long.
Eyes-off technology changes that relationship. The goal is for the automated system, when operating inside its approved conditions, to perform the driving task without requiring continuous visual supervision. GM has described scenarios in which an occupant could check email, work through a to-do list or watch entertainment instead of staring continuously at traffic. That is a much bigger technical and legal step than hands-free driving. Under the SAE framework, Level 2 systems require continuous driver supervision, while higher automation levels shift responsibility for performing the driving task to the automated system within defined conditions.
GM Does Not Want to Call It Conventional Level 3
The interesting wrinkle is that GM increasingly resists describing its system simply as “Level 3.” Chief Product Officer Sterling Anderson has argued that the traditional automation scale can create an awkward middle ground in which a driver is encouraged to disengage mentally but must still be prepared to rescue the system. GM says it wants to avoid building autonomous technology around that kind of sudden emergency handoff.
That position does not make established definitions disappear. SAE defines Level 3 as conditional automation in which the system performs the driving task under defined conditions while a human remains available when an intervention request is issued. Level 4 goes further by requiring the automated system itself to manage fallback within its operating domain. GM’s public language increasingly emphasizes an autonomous highway system that can function without constant supervision rather than focusing on a particular SAE badge. For buyers, the eventual operating rules will matter more than terminology: what conditions are supported, what happens when those conditions end, and exactly when the human must resume driving.
Pickups and Big SUVs Are Where GM Can Reach Real Scale
Moving the technology into trucks and large SUVs dramatically changes its commercial significance. Cadillac’s Escalade IQ is a premium electric vehicle, while Chevrolet Silverado and GMC Sierra pickups sell in volumes that can put new technology in front of hundreds of thousands of customers. During the third quarter of 2026 alone, GM delivered 670,974 vehicles in the United States, and the Silverado and Sierra remained among the company’s highest-volume products. GM also continued to lead the full-size pickup and full-size SUV segments.
The company is already preparing truck buyers for more advanced driver assistance. GM told shareholders that Super Cruise will become standard on high-end versions of the next Silverado and Sierra and optional on most other trims, with Super Cruise penetration across its truck portfolio expected to increase four- to fivefold in 2027. That creates a natural customer pipeline for whatever comes next. A technology introduced through Cadillac can establish credibility; putting it into mainstream work trucks, family haulers and long-distance tow vehicles is what could turn it into a large-scale business.
A New Computer Architecture Makes the Expansion Possible
GM argues that spreading autonomy across different vehicles requires changing the electronics underneath them. Traditional vehicles can contain dozens of separate electronic modules, each responsible for individual functions. GM’s next-generation design consolidates far more of that work into a liquid-cooled Central Compute Unit connected to zonal controllers through high-speed Ethernet. The company says the central computer will offer roughly 40 times the computing capability of today’s GM vehicles, while the new network can move about 36 times as much data as previous systems.
That common foundation is designed for both gasoline and electric vehicles, which is especially important if eyes-off capability is headed to conventional pickups and SUVs rather than being reserved for EVs. GM says software is separated from much of the vehicle-specific hardware, allowing validated improvements to spread across multiple models through over-the-air updates. Its fourth-generation EV platform will also use about 30% less wiring than today’s GM EVs. The autonomous-driving stack adds sophisticated sensing, including high-resolution radar, while GM previously said its eyes-off system would use lidar, radar and cameras for redundancy.
Billions of Virtual Miles Are Joining Real-World Testing
Autonomous systems have to cope with the events engineers expect and the strange ones they do not. GM says Super Cruise customers have now accumulated more than 1.2 billion hands-free miles, giving the company a large base of real-world information about highways, traffic, weather and driver behavior. Nearly 750,000 Super Cruise-equipped vehicles were on the road as GM built that dataset.
The next phase is substantially more aggressive. Hundreds of GM data-collection vehicles are generating roughly one million miles of full-rate, multisensor information every month, according to the company. GM also says its simulation system now produces more than three billion virtual driving miles every day. Earlier in 2026, more than 200 development vehicles were being introduced into supervised public-road testing in California and Michigan. Simulation allows developers to repeatedly create difficult highway merges, unusual traffic behavior and extremely rare hazards without waiting years for them to occur naturally. The company has even demonstrated synthetic scenarios involving a giraffe entering a vehicle’s path—an exaggerated example designed to show how unusual events can be generated and tested safely.
GM Is Testing People as Seriously as It Tests Cars
Building an autonomous system is only part of the problem. Engineers also need to know what happens to people once they are told they no longer have to watch the road. At GM’s technical facilities in Warren, Michigan, researchers have put volunteers inside real vehicles surrounded by simulated environments containing traffic, hills, construction, rain and other conditions. Researchers can observe eye movements and physiological responses while introducing distractions that resemble what occupants could actually do during an eyes-off trip.
One Associated Press demonstration involved a researcher playing a mobile game during simulated autonomous driving. The exercise may sound lighthearted, but the underlying question is serious: how should a vehicle communicate that it is safe to stop paying attention, and how should it bring someone back into the driving task when necessary? GM has previously shown turquoise lighting as a visual signal for autonomous operation. Its research teams have repeatedly emphasized trust, arguing that technically capable automation will not succeed if occupants either fear the system or become confused about what it can actually handle.
GM Is Entering an Increasingly Crowded Autonomy Race
GM will not arrive in an empty market. Mercedes-Benz became the first manufacturer authorized to offer an SAE Level 3 production system in the United States with Drive Pilot, initially gaining approvals in Nevada and California. Its early American operating conditions were deliberately restrictive, including suitable highways, heavy traffic and speeds of no more than 40 mph. That demonstrated that eyes-off consumer automation was possible, but also highlighted how narrow its usable domain could be.
Ford has since announced plans for its own eyes-off technology in 2028, initially tied to its new affordable electric-vehicle platform. Tesla’s Full Self-Driving (Supervised), despite performing lane changes, turns, intersection handling and other complex maneuvers, still officially requires active driver supervision. Ford’s current BlueCruise similarly remains a hands-free, eyes-on Level 2 system. GM therefore has an opportunity to distinguish itself if it can deliver broadly usable highway autonomy at scale. It also faces the same regulatory, liability and human-factors questions that have slowed competitors, with automated-driving rules and permitted use differing across jurisdictions.
The Bigger Prize Is Software Revenue — and Eventually a Car That Runs Errands
GM’s autonomy push also fits a broader effort to earn more money after the vehicle leaves the dealership. The automaker expects to approach $7.5 billion in deferred revenue from software and services during 2026, and Super Cruise is becoming an increasingly important part of that strategy. GM reported earlier this year that 30% to 40% of customers whose three-year prepaid Super Cruise period had ended were attaching a paid plan. Expanding advanced automation into popular trucks could dramatically enlarge the potential subscriber base.
Still, several crucial details remain unanswered. GM has not announced what eyes-off driving will cost, whether access will require a recurring subscription, which pickup or full-size SUV will receive it first, or when the expansion beyond Cadillac will begin. The company’s longer-term ambition goes even further: vehicles capable of leaving without their owners to pick up groceries, collect dry cleaning, move family members or park themselves. That “personal autonomy” vision remains farther away. The more immediate test comes in 2028, when GM has to prove that an Escalade IQ can safely give its driver something automakers have promised for years but rarely delivered: meaningful time back.
