General Motors is still very much in the business of selling gas-guzzling trucks and massive SUVs. But if you listen to its leadership, the Detroit giant is rapidly transforming into a nimble, software-defined tech player with a tightened grip on AI. That was the pitch today, when GM hosted its tech-focused Onward event to unveil a wide-ranging slate of EV, autonomous vehicle, and AI initiatives — all aimed at fending off the rising threat of tech-heavy Chinese automakers.
The promises are big: GM claims it will use AI to cut vehicle development times in half, introduce compute units up to 40 times more powerful than its current hardware, model next-gen battery chemistries in record time, and bring eyes-off autonomous driving to its top-selling trucks and SUVs. In short, GM wants to transition from a legacy automaker into an AI-driven tech enterprise, leveraging its massive scale across both electric and gas-powered fleets. It’s a bold strategy, even if we’ve heard variations of it before.
[Image: https://platform.theverge.com/wp-content/uploads/sites/2/2026/10/Pre-production-vehicle-at-PPO.jpg?quality=90&strip=all] Faster R&D and digital factory twins For decades, GM has wrestled with sluggish four- to five-year vehicle development cycles that lag behind the light-speed pace of consumer tech. The company tried to bridge that divide with aggressive EV targets and software like Super Cruise.
But as software-defined architecture evolved from an industry buzzword into a table-stakes requirement, maintaining fragmented vehicle platforms and dozens of disparate control units began to severely drag down engineering efforts. Now, GM says it’s aiming to slice that development cycle down to just two years using “digital thread virtualization.” Instead of operating in silos or relying on slow, step-by-step handoffs, the company says it will unify design, engineering, and manufacturing digitally from day one.
(We’ve already reported on GM’s efforts to inject AI directly into its vehicle design process.) GM points to its new battery development program as proof of concept. The automaker tested over 300,000 virtual battery designs before building a single physical prototype, eventually settling on a lithium manganese rich (LMR) chemistry. That virtual process let GM’s battery team go from initial concept to formal approval in just one month. These new LMR cells will power GM’s fourth-generation EV platform.
The company claims the cells offer 33 percent higher energy density than traditional lithium iron phosphate cells at a comparable cost alongside a 20 percent boost in efficiency per kilowatt-hour, yielding longer-range EVs without forcing drivers to haul around a heavier pack. The batteries will be produced by Ultium Cells, GM’s joint venture with LG Energy Solution, at its factory in Spring Hill, Tennessee. Beyond the battery, GM engineered the entire powertrain as a single integrated system.
The new drive units feature 40 percent fewer internal parts than current hardware. Because the drivetrain is lighter and takes up significantly less space, future cabins theoretically should offer more room for passengers and cargo.
[Image: https://platform.theverge.com/wp-content/uploads/sites/2/2026/10/CCU-installed-in-pre-production-vehicle-2.jpg?quality=90&strip=all] A single, liquid-cooled brain To fix its messy electronic setup, GM is replacing dozens of small, scattered electronic control units and heavy wiring harnesses with a single central computer. GM claims this unified unit — which is liquid-cooled and connected by high-speed lines — delivers 40 times the computing power while stripping out 30 percent of the physical wiring.
That extra horsepower should translate into snappier screens, smarter voice controls, and seamless over-the-air (OTA) updates for automated driving features and battery management. Granted, GM has been promising smarter cars and faster software updates for years. The automaker launched a “digital nerve system” in the 2020 Cadillac CT5 and later rolled out its Ultifi software platform with big promises about OTA capabilities.
While drivers generally give GM high marks for hardware dependability, the company’s software rollout has faced its fair share of stumbles and consumer frustration over unnecessary complexity.
(Remember the disastrous Chevy Blazer rollout?) GM is extending this digital integration to the factory floor with a project dubbed — and capitalized rather oddly — “PlantBrAIn.” By pairing real-time factory floor data with digital-twin simulations across stamping, casting, welding, painting, and assembly, GM hopes workers can catch physical errors and equipment bottlenecks in a virtual model before they disrupt actual production.
[Image: https://platform.theverge.com/wp-content/uploads/sites/2/2026/10/AV-Research-Vehicle1.jpg?quality=90&strip=all] Personal autonomy — and an autonomous dry-cleaning pickup? GM’s newly upgraded compute platform will make its production debut in 2028 on the Cadillac Escalade IQ — the same vehicle GM announced last year would be the first to receive its hands-free, eyes-off highway driving system.
(Given that GM teased this centralized compute platform last year, today’s hardware announcement is a bit of a retread.) However, GM isn’t keeping its eyes-off tech locked behind the price tag of a $120,000 luxury SUV. The automaker confirmed plans to roll out hands-free, eyes-off highway driving across its broader lineup, spanning both EVs and gas-powered models. As GM puts it: > Our vision is to build vehicles that can drive themselves or their occupants anywhere they need to be.
This includes a future where your vehicle can handle mundane tasks for you, like heading out into bumper-to-bumper traffic to pick up groceries or dry cleaning, giving you valuable time back. An uncrewed pickup truck executing a dry-cleaning run through morning traffic is certainly a vision, though one we’ll need to see to believe. GM also teased a futuristic vision for autonomous cabins called “Concept Horizon,” which reimagines the interior as an open space purely dedicated to relaxing, working, or unwinding.
[Image: https://platform.theverge.com/wp-content/uploads/sites/2/2026/10/OnStar-Mobile-App-stills2.png?quality=90&strip=all] Overhauling OnStar for its 30th birthday Finally, GM is celebrating OnStar’s 30th anniversary by giving the iconic service a massive AI overhaul. Rather than acting strictly as a backup button for roadside emergencies, OnStar is being rebranded as an active, AI-driven intelligence platform.
Drivers can trigger the new OnStar AI Assistant by saying, “Hey OnStar” to tweak cabin temperatures, open the glove box, pop the EV charge port, or check tire pressure and battery diagnostics. The assistant will debut on the 2027 Cadillac Lyriq through an OTA update before expanding to other models. While GM previously rolled out Google’s Gemini-powered voice assistant to 2022 and newer models, this new OnStar assistant operates at a deeper system level.
Unlike Gemini, it can execute physical vehicle functions and run hardware-level diagnostics, such as managing specific HVAC zones, interpreting dashboard warning lights, or running battery health checks on the fly. GM’s vision of a central, AI-driven vehicle architecture is undeniably ambitious, and if it works, it could genuinely streamline how cars are built and updated. But for an automaker that has spent years pitching software transformations while occasionally stumbling, execution remains everything.
Until an uncrewed Silverado actually pulls into a suburban driveway with a load of fresh dry cleaning, drivers will likely treat these grand AI promises like any other product roadmap: best taken with a healthy dose of patience.
منبع: ورگ
