GM lays off 600 IT workers in AI skills swap as automaker pivots to software-defined vehicles

General Motors is laying off 500 to 600 salaried IT workers and replacing them with engineers who know how to build AI systems. The cuts, centred in Austin, Texas, and Warren, Michigan, represent more than 10 per cent of GM’s IT department. The company is not reducing headcount for cost savings. It is executing what amounts to a skills swap: clearing out workers whose expertise no longer fits and hiring data engineers, prompt specialists, and AI-native developers in their place. GM shares fell four per cent after the announcement.

The restructuring is narrow. It targets one department in a company that employs roughly 163,000 people. But its significance extends beyond the numbers because it illustrates what enterprise AI adoption looks like in practice at a 118-year-old industrial company. GM is not bolting AI tools onto its existing workforce. It is deliberately rebuilding the workforce from the ground up, role by role, to match the technical demands of a product strategy that has shifted dramatically in the past 18 months.

The strategy

GM bet on three technology futures: electric vehicles, autonomous driving, and software-defined vehicles. It is now retreating from two and doubling down on the third.

The electric vehicle pullback has been severe. GM recorded 7.1 billion dollars in special charges in the fourth quarter of 2025, including six billion related to its EV plans. It sold its stake in the Ultium Cells battery plant in Lansing, Michigan, to LG Energy Solution. It idled battery plants in Ohio and Tennessee. It laid off 1,750 workers indefinitely and temporarily cut another 1,670 at EV and battery facilities. In April, GM suspended development of its next-generation full-size electric truck and SUV programme entirely, affecting refreshed versions of the Chevrolet Silverado EV, GMC Sierra EV, GMC Hummer EV, and Cadillac Escalade IQ.

The autonomous driving retreat came earlier. In December 2024, GM shut down Cruise’s robotaxi operations after years of investment and a series of safety incidents. It absorbed the Cruise technical team into GM’s own autonomous vehicle division and redirected the effort toward personal vehicles rather than ride-hailing. Nearly 90 per cent of the code for GM’s autonomous driving software is now written by AI, a figure that would have been unthinkable two years ago and that partly explains why the company needs fewer legacy IT workers and more engineers who can manage AI-generated code at scale.

The software-defined vehicle strategy is what remains. GM is integrating Google’s Gemini conversational AI into its vehicles starting this year. It has contracted with Nvidia to replace the current Qualcomm Snapdragon platform with Nvidia’s Drive Thor system, which will debut in 2028 powering an entirely new electrical and electronic architecture designed for software-first vehicles. The company has hired Behrad Toghi from Apple as its AI lead and Rashed Haq, formerly of Cruise, as vice president of autonomous vehicles.

The earnings

The IT layoffs come after a strong first quarter. GM reported 43.6 billion dollars in revenue, roughly flat year over year. Adjusted EBIT was 4.3 billion dollars, up 22 per cent. Adjusted earnings per share were 3.70 dollars, beating Wall Street estimates by 40 per cent. Net income was 2.6 billion dollars.

The quarter included a 500 million dollar benefit from a US Supreme Court decision that terminated and refunded certain tariff levies paid under the Trump administration’s trade policies. GM raised its full-year 2026 adjusted earnings guidance by 500 million dollars to 13.5 to 15.5 billion, reflecting the tariff rebate. It also booked 1.1 billion dollars in special charges related to the EV pullback, with 90 per cent of expected supplier claims now recorded and most cash outflows expected to conclude this year.

The company is profitable, growing earnings, and generating strong cash flow from its North American truck and SUV business, which accounts for the vast majority of its operating income. The IT restructuring is not a distress signal. It is a company that can afford to rebuild its technology workforce choosing to do so while the core business is healthy.

The pattern

Zuckerberg told Meta employees the layoffs are about capital expenditure, not AI productivity, a framing that mirrors GM’s approach. Both companies insist the cuts are investments, not austerity. Both are spending the savings on AI infrastructure. Both face the same credibility question: if the cuts are really about building the future, why do the workers being cut learn about it the same way workers learn about cost cuts?

Meta is cutting 8,000 jobs while spending 135 billion dollars on AI. The scale differs from GM’s 600 IT positions, but the logic is identical. Companies across industries are concluding that their current workforces were assembled for a pre-AI era and that maintaining them is an opportunity cost. The money spent on salaries for workers with legacy skills could be spent on workers with AI skills or on the AI systems themselves. The restructuring is the reallocation.

Atlassian launched AI visual tools and partner agents in Confluence one month after cutting 1,600 jobs. The developer tools company followed the same sequence as GM: cut workers, announce AI investment, ship AI features. The pattern has become so consistent across industries that the question is no longer whether companies will restructure around AI but whether the restructurings will produce the outcomes they promise.

The workforce question

Chinese courts have ruled that AI replacement is not legal grounds for firing workers, a decision that draws a line the United States has not. In China, an employer cannot terminate an employee solely because AI can perform the job. In the US, GM can lay off 600 IT workers, announce it will hire AI-native replacements, and describe the process as transformation rather than termination. The legal and cultural frameworks for AI-driven workforce restructuring are diverging globally, and GM’s approach sits squarely on the American side of that divide.

The skills swap model assumes that the workers being hired are available, that they can be integrated into a 118-year-old manufacturing company’s culture, and that the AI systems they are building will deliver the productivity gains that justify the disruption. None of these assumptions are guaranteed. The cybersecurity, software engineering, and AI talent markets are among the most competitive in the world. GM is competing for the same workers as Google, Nvidia, Apple, and every other company executing the same pivot. An automaker in Warren, Michigan, is not the obvious first choice for a machine learning engineer choosing between offers.

The position

The argument that AI’s ability to replace jobs should not be flaunted captures the tension in GM’s announcement. The company is not hiding the fact that it is replacing workers with AI-skilled ones. It is presenting it as strategy. The IT layoffs are framed as an upgrade, not a reduction. The language of transformation, optimisation, and skills alignment is designed to distinguish this from a cost cut.

But the workers being laid off in Austin and Warren are not being transformed. They are being replaced. The distinction between a cost cut and a skills swap matters to investors and executives. It does not matter to the person who just lost their job because the company decided it needs a different kind of engineer. GM is a profitable company that posted a 40 per cent earnings beat, raised its guidance, and then announced it is firing 600 people because their skills are the wrong skills for a world in which 90 per cent of autonomous driving code is written by AI. The strategy may be correct. The execution may work. The workforce question, what happens to the workers whose skills were right until the company decided they were not, is the one that no restructuring memo answers.