General Motors announced on Tuesday it will end robotaxi development at its money-losing Cruise business, a blow to the ambitions of the largest US automaker to advance the technology.

GM said it would no longer fund work on self-driving robotaxis “given the considerable time and resources that would be needed to scale the business, along with an increasingly competitive robotaxi market”.

Instead, GM will prioritize developing Super Cruise, its advanced driver assistance system for personal vehicles, and Cruise will be folded into its group working on driver assistance technology. Mary Barra, the GM CEO, declined to say how many Cruise employees could be moved over to GM.

The development marked a significant shift for GM. The automaker has invested more than $10bn in Cruise since 2016. Just last year, Barra said the Cruise business could generate $50bn in annual revenue by 2030.

But on Tuesday, Barra said the business was expendable. “You’ve got to really understand the cost of running a robotaxi fleet, which is fairly significant, and again, not our core business,” the CEO said on an analyst call.

Marc Whitten, the Cruise CEO, who has been in the role since June, said that Cruise’s board of directors and its leadership team are “working closely” with GM on next steps.

Kyle Vogt, the founder and former CEO of Cruise and overall champion of self-driving technology, expressed frustration with GM’s decision. “In case it was unclear before, it is clear now: GM are a bunch of dummies,” Vogt posted on X.

Costly business

GM’s decision comes as it has scaled back plans for electric vehicles, selling its stake in one of its joint venture battery plants and restructuring its China business, to focus more on its profitable business of making gasoline-powered pickup trucks and other large vehicles.

But it’s not the first in the emerging robotaxi industry to scale back ambitions. Ford Motor has wound down its Argo AI operation, which was partly funded by Volkswagen. The company is still working on advanced driver assistance systems in-house. Uber and Lyft also invested heavily in driverless car technology systems they have since shuttered.

The cuts highlight the harsh reality facing others still in the race: it requires a long-term commitment to perfect the technology and deep pockets to fund it.

“The decision from GM raises an interesting question of whether AV economics can work at all,” Bernstein analysts said in a note. “They can, but it requires capable tech and a willingness to spend billions if an AV provider is keen to scale a proprietary network, as we saw in the early days of rideshare.”

Left in the field are developers like Alphabet’s Waymo, the only company that runs paid, unmanned taxis in the US; Tesla, led by the billionaire Elon Musk, a close Trump adviser; and Amazon.com’s Zoox, which is testing a vehicle that has no manual driver controls such as a steering wheel and pedals. Chinese companies, including Baidu’s Apollo and WeRide, are also testing autonomous vehicles in the US.

Musk is bullish on the future of robotaxis, even more so amid his deepening ties with Trump.

Waymo last week said it would expand its autonomous ride-hailing services to Miami. Last month, the company opened its ride-hailing services to everyone in Los Angeles, and in October it closed a $5.6bn funding round led by Alphabet.

Headwinds

With nearly $10bn from GM, Cruise had launched commercial operations last year and was once considered an industry frontrunner, but it remained a money-losing enterprise.

Ultimately, it was unable to recover from a 2023 accident in San Francisco when one of its self-driving vehicles dragged a pedestrian 20ft and then stopped on top of her, leaving her critically injured.

Cruise provided video footage of the crash to the Department of Motor Vehicles, which the department said omitted key aspects of the incident. When the DMV got the video 10 days later, it ordered Cruise to immediately halt all operations in California.

In a cascade of events, Cruise recalled and grounded its entire fleet of vehicles and faced state and federal government investigations and fines. Vogt, then CEO, resigned and nearly a dozen other Cruise executives stepped down.

At the time of Vogt’s departure, Barra wrote in an email to employees that she and the board were “intensely focused on setting up Cruise for long-term success”. The priority, she wrote at the time, was regaining public trust and accountability would be a big part of that.

An investigation by the justice department said Cruise failed to disclose key details of the crash to regulators. GM paid a substantial settlement to the woman who was injured.

In May, Cruise resumed supervised autonomous driving in Phoenix with safety drivers in a bid to make a comeback. But in July, GM said it would halt development of a planned robotaxi that would not have a steering wheel or other human controls.

Ultimately, Cruise admitted to submitting a false report to influence a federal investigation under the jurisdiction of the National Highway Traffic Safety Administration and agreed to pay a $500,000 criminal fine as part of a deferred prosecution agreement.

For others still developing or operating robotaxis, Cruise’s exit sends a clear warning, Philip Koopman, a Carnegie Mellon University professor working on autonomous vehicle safety, told Reuters.

“The cost of having a bad crash, especially where it looks like you’re not paying enough attention to safety as you should have been, could be the whole company,” he said.

“That’s a reason to be mindful of safety even as you’re under pressure from investors to make quick progress.”