GM is betting compute and AI will assist it compete with Chinese automakers

When GM introduced its new fourth-generation EV platform on Wednesday, what the automaker didn’t share was nearly as notable as what it did.

The new EV platform, a part of a collection of technology-focused introduced throughout its annual GM Forward occasion in New York, didn’t discuss suspension geometries, dash-to-axle ratios, and torsional rigidity, all nerdy automotive phrases that translate to how individuals will expertise a automotive. Is it supple or sporty? Does it have a refined stance? When it hits a pothole, does it really feel like a financial institution vault or a paper plate?

Instead, GM’s announcement centered on novel battery chemistries, revamped electrical motors, and a completely new central computing platform and electrical structure that may join each system within the automotive, together with acceleration and energy, steering, braking, and lively security options in addition to software program. Lots and plenty of software program.

“We suppose automobiles can change into a extra succesful and useful a part of on a regular basis life if we make them areas that perceive context, scale back friction, and provides individuals again moments that will in any other case be consumed by the calls for of the highway,” Sterling Anderson, government vp, international product, and chief product officer at GM, informed TechCrunch in a latest interview.

That positive doesn’t sound like a conventional automaker’s pitch. Then once more, for the reason that arrival of Tesla, Rivian, and Chinese automakers, the business hasn’t been the identical. Ride and dealing with nonetheless matter, however they’re desk stakes. What units automobiles aside at present is essentially outlined by software program.

Anderson has talked about this future GM earlier than. Last yr, the corporate introduced plans to overtake {the electrical} and computational guts of its future automobiles in a bid to ship quicker software program, extra succesful automated driving options, and a customized, conversational AI assistant.

This yr, GM is offering extra particulars and dates. This technological overhaul is about to debut in 2028 with the Cadillac Escalade IQ. The firm will even rollout an autonomous highway driving feature in 2028 that may enable drivers to have their arms and eyes off the highway and with out the have to be able to take over.

The software program race consists of {hardware}

European and U.S. automakers have spent years chasing the prospect of software-defined automobiles, however have typically fallen brief. Volkswagen Group had a lot bother with its personal in-house effort, together with its troubled Cariad unit, that it struck a $5.8 billion joint venture deal with relative newcomer Rivian.

GM has been by itself capitally intensive software program journey. The firm has guess closely on software program in recent times. It ditched the favored smartphone projection know-how Apple CarPlay for practically all of its EVs, and as an alternative turned to a Google-based infotainment system, which is definitely fairly succesful even when it doesn’t but reside as much as the requirements set by Tesla and Rivian.

It’s poured billions into software program and autonomous car know-how improvement too. GM’s capitalized software program amortization expense in 2025 was $866 million, which supplies a way of how a lot it’s spending even because it spreads out the associated fee over years.

Like many legacy automakers, GM has come to the conclusion that the outdated method to electronics, which scattered pc chips all through the car, was holding it again. 

For fourth-generation EVs, GM will centralize a lot of these features on a Central Compute Unit (CCU) that’s 40-times extra highly effective than its present techniques. Eliminating smaller chips on the perimeters will assist trim wiring by 30%, saving weight and simplifying meeting.

GM notes that this new CCU will work on fossil gasoline fashions, too, although it should debut within the Cadillac Escalade IQ, a car that’s shaping as much as be the automaker’s technological flagship.

The CCU will deal with the same old propulsion and chassis controls, but additionally notion and lively security. That ought to assist pave the best way for extra superior autonomy options, which GM additionally introduced on Wednesday.

Such a transfer was maybe inevitable. Pressure on automakers has been constructing on all sides. 

Domestically, Tesla has been pushing the boundaries of autonomy, partly as a result of it developed a real software-defined car. 

Automakers like GM are additionally feeling stress from Chinese producers, which have shortened improvement cycles to 2 to 3 years, down from the same old 4 to 5 years.

GM intends to match that cadence. Alongside bulletins about battery chemistries and autonomy, GM outlined how enterprise AI instruments have sped up car improvement throughout each sector because it pushes to slash its conventional four-to-five-year improvement timeline in half.

GM gave a short glimpse of its revamped improvement technique. One instance: engineers went by means of 300,000 digital designs of a brand new lithium-manganese-rich battery pack earlier than they assembled the primary bodily prototype. That helped the idea transfer to an authorised program in a single month.

Such velocity will assist GM save prices because it races to protect profitability amid a historic upheaval within the automotive business. But will it’s sufficient to catapult it again into the management place it as soon as loved?

The rise of Chinese automakers received’t make it simple. Their factories are able to producing excess of Chinese customers can soak up, and paired with authorities subsidies and cheaper labor, these automobiles have flooded financial hubs from Europe to South America with cheap automobiles. 

Chinese automaker’s price benefit is so steep that, with out commerce limitations, “they may just about demolish most different corporations within the planet,” Elon Musk said in a 2024 earnings name. In response, he has been making an attempt to pivot Tesla from an automotive firm to an AI and robotics firm.

GM’s transfer into autonomy indicators an analogous shift. But to ship true autonomy, GM realized it didn’t have the platform it wanted. With a brand new structure, it ought to have a combating likelihood.