So it came as a surprise on Tuesday when General Motors so suddenly changed its tune—announcing that it would stop funding Cruise’s robotaxi service and would bring the startup’s technical employees in-house, where they’ll focus on self-driving technology for the autos GM sells directly to consumers.

In a call with analysts, GM CEO Mary Barra cited the “time and expense” that would have been required to keep scaling the robotaxi business. She also cited increasing competition, even though Cruise really only has one major competitor these days—Alphabet’s Waymo, which offers its robotaxi service in Los Angeles, Phoenix, and San Francisco, and will soon debut in Miami, Austin, and Atlanta.

But it’s difficult to ignore the high profile 2023 accident in San Francisco that had led GM to temporarily halt its service. The incident, which sidelined Cruise operations for more than a year, underscores the treacherous terrain facing companies seeking to lead the way in the nascent business of robotaxi services. 

For automakers in particular, the combination of developing complex self-driving technology and managing a robotaxi fleet on public roads—with all the inherent risks involved—has proven to be a humbling journey. General Motors now joins automakers like Ford Motors and Volkswagen who also once made big, billion-dollar bets on the future of ridehail, only to back away from those plans not long after. 

That leaves Tesla as the only American pure-play automaker still competing in the robotaxi race, alongside internet companies like Alphabet and Amazon and several Chinese competitors. But the move away from robotaxis also raises questions about the future of traditional automobile makers like GM, as they seek instead to make self-driving technology a feature in privately owned cars—a shift that could dent their business models. 

‘A lot of uncertainty’

Cruise was supposed to be the shiny new piece of technology that could reposition $57 billion General Motors from a legacy automaker into a true, innovative Tesla rival for the auto industry’s next chapter. And for a while, it was. Cruise was the first company to launch a paid fare operation in San Francisco, and, at one point, it had 4,000 employees in Seattle, Phoenix, Austin, Dubai, and in Japan.

But it was never going to be cheap. General Motors spent $1 billion to acquire the less-than-50-person team in 2016, and went on to spend more than $10 billion on the company over the next nine years. That’s because, while robotaxis may be self-driving, they require an army of people to operate a service—including scores of contract workers to manage and test the cars, address damage, or remotely assist the vehicles if they get stuck. 

It’s these steep costs, in part, that have led backers to walk away or companies to shut down. 

Safety concerns have added another challenge.

In 2018, an Uber struck a woman on a bike in Arizona, killing her. It was the first fatality from an autonomous software system, and Uber never managed to regain trust after it. Uber sold its self-driving business two years later as it focused on trying to get profitable.

Cruise never quite recovered from its own high-profile accident last October, when one of its robotaxis ran over, then dragged a woman for 20 feet beneath it. Regulators accused the company of withholding information about the incident, and Cruise parted ways with almost its entire executive team over the incident after grounding its fleet.

While Cruise hired a new batch of executives and was in the process of working through the plan it laid out to restart operations, the business continued to cost the Detroit-based carmaker $1.3 billion this year, with little revenue coming back in. And with more than $10 billion in sunk costs, a return on investment still seemed pretty far off—particularly as rival Waymo continued to expand into new cities in the U.S.

“I think Cruise would have been very capable of building up a large fleet and pulling in revenue, but to get to profitability and really add to GM’s bottom line—it’s just way out there with a lot of uncertainty,” says Richard Bishop, a consultant who publishes regular reports on the robotaxi industry.

General Motors’ shift away from the robotaxi business now leaves an empty space for its $1.3 trillion rival Tesla, which has been talking about ramping up plans for its own robotaxi business in the future and has started conducting limited testing in San Francisco.

“It’ll be a great talking point to make them look better,” Bishop says of Tesla.

It’s still unclear exactly how Tesla plans to compete—and the operational approach it will take with its so-called cybercab—but Tesla has now become the last of the American carmakers still dabbling in the robotaxi business, alongside Waymo and Amazon’s Zoox and several Chinese companies. 

The Road Ahead

GM’s new plan to integrate self-driving technology into its own cars, meanwhile, looks a lot like Tesla’s existing, and controversial, approach of offering limited autonomous driving capabilities as an add-on feature in its electric vehicles. Federal regulators are looking at a series of accidents involving Teslas and its so-called Autopilot and Full Self-Driving features, which the company says require active driver supervision.

In fact, GM already offers a limited set of autonomous driving features in some of its autos through “super cruise,” which can steer vehicles like the Hummer EV on highways (while the driver pays close attention behind the steering wheel). The leap to the next level of autonomous driving is far more significant though, and brings a step increase in costs and risks. 

Do car buyers even want to own a fully autonomous car, or is the joy of driving what gives a car its real appeal and value? And, assuming GM can make Cruise’s technology safe and affordable enough to integrate into its cars, will they be able to compete in a world where passengers can get to their destination by hopping in a robotaxi? 

Barra and GM may have escaped the turbulence and costs of its Cruise robotaxi episode, but its self-driving car problem has not gone away.