Why China Is Pumping the Brakes on Robotaxis to Protect Driver Jobs

China has spent years building the world's largest robotaxi industry, but in 2026 the government is deliberately slowing it down to protect the millions of people who drive for a living. According to reporting from Bloomberg, Beijing now faces a genuine dilemma, which is how to keep its lead in autonomous-vehicle technology without throwing taxi and ride-hailing drivers out of work. At Metaintro, we track how AI policy reshapes real jobs, and this story is a preview of a fight that drivers, couriers, and gig workers everywhere will eventually face.

The tension came to a head this spring. On March 31, 2026, dozens of driverless cars operated by Apollo Go, the robotaxi arm of search giant Baidu, suddenly froze in the middle of busy roads in the central megacity of Wuhan. The mass outage snarled traffic and went viral on Chinese social media within hours. Within weeks, regulators stopped issuing new licenses for autonomous vehicles nationwide, and as of early June they still had not resumed. People familiar with the review told reporters it could take a while before new permits flow again, because authorities want a full accounting of what went wrong before they let the fleets keep growing.

What makes the moment unusual is that the slowdown is not really about that one glitch. China could have lifted the freeze quickly and called it a software patch. Instead, officials are using the pause to step back and ask a harder question, which is whether a country that already leads the world in autonomous driving wants to push that lead at the cost of millions of paychecks. That is the dilemma Bloomberg described, and it is one almost no government has been willing to confront so openly.

Why Did China Hit Pause on Robotaxis?

The licensing freeze was triggered by the safety scare, but it sits on top of a deeper worry about jobs. This is not the first pause. Chinese authorities also suspended licensing in 2024 after complaints flared from residents and taxi drivers in Wuhan over the very same Apollo Go service, a rollout that sparked heated public debate at the time. That earlier episode set the template, which is that a public incident gives officials cover to slow a technology they already feel uneasy about.

After the March outage, the Chinese Ministry of Transport and the Ministry of Industry and Information Technology convened a meeting in early April with eight robotaxi companies and officials from cities running pilot programs, as the trade outlet Electrive reported. Providers and local governments were told to improve safety measures before any new licenses would be granted. Existing services may keep running, but companies cannot add cars, launch new test projects, or expand into new cities while the freeze holds.

The 2024 precedent is worth dwelling on, because it shows the jobs angle is not new. That year, a local Wuhan taxi operator wrote to the municipal transport authority complaining that several of its cars had stopped working because of declining income, and accused the robotaxis of taking jobs from the grassroots. Residents filed hundreds of complaints on a government-run transport site, and the rollout made both friends and enemies on Wuhan's streets. The pattern repeats, which is that public frustration with driverless cars in China tends to fuse safety anxiety with economic anxiety, and regulators respond to both at once.

How Many Driver Jobs Are Actually at Stake?

The numbers explain why Beijing is treading carefully. By 2024, the count of licensed ride-hailing drivers in China had reached 7.48 million, a 159 percent jump since 2020, according to research cited by Beijing Scroll from the China Research Center on New Forms of Employment. That figure is roughly the entire population of Hong Kong, and it does not include traditional metered-taxi drivers, who number in the millions more.

Even before robotaxis arrived at scale, that workforce was under strain. The supply of drivers had ballooned, while monthly average orders rose just 38.3 percent over the same stretch, as Sixth Tone has documented. Price wars between platforms pushed fares down, forcing drivers to work longer hours just to hit their income targets. Several cities, including Changsha, Dongguan, and Sanya, had already issued saturation warnings and frozen new ride-hailing permits. Robotaxis land on a market that was close to oversupply to begin with.

What Does the Research Say About Driver Income?

A peer-reviewed study published in 2026 in the journal Humanities and Social Sciences Communications, part of the Nature portfolio, put a number on the damage. Researchers found that the arrival of robotaxis cut traditional taxi drivers' average daily income by 10.9 percent, largely because riders shifted to the cheaper driverless option. The same study reported that robotaxis pushed drivers to work longer hours, raised their job stress, lowered job satisfaction, and made many of them start looking for other work.

That last finding matters most for job seekers. A 10.9 percent income hit is the kind of squeeze that quietly forces a career change. Workers do not always wait to be laid off by a machine. When the math stops working, they leave on their own, often without a clear next step lined up. That is the slow-motion version of automation displacement, and it is already underway in China's biggest cities.

It also reframes how to think about AI and jobs more broadly. The popular fear is a single dramatic moment when a robot takes your seat. The reality the data shows is gentler and more corrosive, which is a steady erosion of earnings that makes a once-viable job no longer pay the bills. By the time the formal headcount drops, many workers have already moved on. For anyone whose role overlaps with what automation can do, that distinction is the difference between reacting early and getting caught flat-footed.

Which Companies Are Driving the Robotaxi Boom?

Three names dominate the field, and all three are expanding fast. Apollo Go has deployed more than 1,000 vehicles across 22 Chinese cities and now completes more autonomous rides in a week than some Western rivals manage in a month. Pony.ai has pushed its robotaxi fleet past 1,700 units and raised its year-end 2026 target above 3,500, while aiming to operate in more than 20 cities by the close of the year, figures the company disclosed in filings to the US Securities and Exchange Commission. WeRide runs more than 800 robotaxis in China, covering over 1,000 square kilometers of Beijing and Guangzhou, according to its own SEC disclosures.

The growth math is what makes the jobs question urgent. A single robotaxi can run almost around the clock without a wage, and once the upfront cost is absorbed, each trip undercuts a human driver on price. That is exactly the dynamic the Nature study captured, and it is why a ride-hailing labor union leader in Guangzhou has publicly called for strict caps on robotaxi fleets, arguing that current rules do too little to safeguard drivers' livelihoods.

Stack those fleet targets together and the trajectory is steep. If Pony.ai alone reaches more than 3,500 cars and 20-plus cities by year end, and Apollo Go and WeRide keep pace, China could have many thousands of driverless taxis competing directly with human drivers within a single year, assuming the licensing freeze lifts. Each of those vehicles represents trips that would otherwise have gone to a person earning a living wage. The companies frame this as progress and convenience, and for riders it often is, with lower fares and short wait times. The friction is that the savings come straight out of driver income, which is the exact tension regulators are now trying to manage rather than ignore.

How Is China Trying to Soften the Blow?

Rather than banning the technology, Beijing is trying to spread the pain over time. The emerging approach blends industrial policy, active labor-market measures, and tweaks to the social safety net so that automation does not translate into sudden unrest. Officials want robotaxi growth to happen in stages, with safety reviews and city-by-city permits acting as natural brakes, while displaced drivers get a runway to retrain or shift into adjacent work.

The city-level patchwork is doing a lot of the heavy lifting here. Because permits are granted locally, individual governments can throttle how many robotaxis operate in their jurisdiction, effectively rationing the disruption to match how fast their labor market can absorb it. Some cities that had already frozen new ride-hailing permits because of oversupply now have a second reason to move slowly on driverless cars. The freeze on autonomous-vehicle licenses simply formalizes that caution at the national level, turning what had been scattered local hesitation into a coordinated pause.

That is a meaningfully different playbook from the hands-off approach in much of the United States, where federal safety rules for autonomous vehicles remain thin and rollouts move faster. China's bet is that managed, slower deployment buys social stability without surrendering its technological lead. Whether that balance holds is the open question, and the rest of the world is watching because the same trade-off is coming for delivery, trucking, and warehouse work everywhere.

There is a hard tension baked into this strategy. Slow the technology too much and China risks ceding ground to faster-moving competitors abroad, undercutting the very industrial leadership it has spent a decade building. Move too fast and it risks the kind of grassroots backlash that already surfaced in Wuhan twice. The current freeze is essentially Beijing buying time to find the middle, and the policy levers it lands on, whether retraining funds, fleet caps, or staged city rollouts, will become a reference point for other governments staring down the same question.

What Does This Mean for Your Career?

If you drive, deliver, or work any job where a machine can plausibly do the task, the lesson from China is to read the early signals rather than the headlines. Robotaxis did not eliminate driving overnight. They chipped away at income first, by about 11 percent on average, which pushed drivers to either grind harder or quietly exit. The smart move is to treat that income squeeze, not the layoff notice, as the real warning, and to start preparing your next move before the math turns against you.

Practically, that means building skills that are harder to automate and that travel across roles. Customer-facing work, hands-on trades, coordination, and anything requiring judgment in messy real-world conditions remain durable. It also means watching policy, because in places like China, the speed of disruption is partly a political choice, not just a technical one. The drivers who fare best will be the ones who start their pivot while they still have leverage, instead of waiting for the fleet to scale past them.

It is also worth borrowing China's framing for your own situation, even if you live far from any robotaxi. Ask which parts of your day are routine enough that a system could plausibly handle them, and which parts depend on relationships, trust, or on-the-spot problem solving. Then steer your time toward the second category. For a driver, that might mean moving into dispatch, logistics coordination, fleet maintenance, or specialized transport that still needs a human. The point is not to outrun automation, which is rarely possible, but to position yourself where it needs you rather than replaces you.

The bigger takeaway is that automation rarely arrives as a clean replacement. It arrives as pressure, on wages, on hours, and on bargaining power, long before any single job formally disappears. China is one of the first governments to admit that out loud and act on it. For workers anywhere whose livelihood touches a steering wheel, a delivery bag, or a forklift, the robotaxi story is less a distant headline and more a dress rehearsal.

People Also Asked

Q: Why is China slowing down robotaxis in 2026?

A: China froze new autonomous-vehicle licenses after a March 2026 incident in which dozens of Apollo Go robotaxis stalled in Wuhan, but the deeper reason is jobs. Officials are weighing the country's AI leadership against the income of millions of taxi and ride-hailing drivers, and the pause buys time to manage that trade-off.

Q: How much do robotaxis reduce taxi driver income?

A: A peer-reviewed 2026 study in Humanities and Social Sciences Communications found that robotaxis cut traditional taxi drivers' average daily income by 10.9 percent, as riders shifted to cheaper driverless trips. The study also reported longer working hours, higher job stress, and more drivers seeking other work.

Q: How many driver jobs could robotaxis affect in China?

A: China had roughly 7.48 million licensed ride-hailing drivers by 2024, up 159 percent since 2020, plus millions of traditional taxi drivers. That large and already strained workforce is why regulators are staging robotaxi growth carefully instead of letting fleets expand all at once.

Future-proof your career with Metaintro, AI-powered tools that help you adapt and thrive. Sign up to stay ahead.