China's Robotaxi Push and the Workers in Its Path

In Beijing's Yizhuang district, driverless vehicles have become a common sight. Robotaxis weave through traffic alongside ordinary cars, while autonomous delivery vans glide along the inside lane as they carry packages to collection points, according to BBC News. For residents, this has apparently passed the point of novelty. The cars are just there now, part of the street's texture.

That normalization is, depending on your position in the economy, either the reassuring proof-of-concept the industry has been waiting for, or the thing that should be keeping you up at night.

How the EV advantage translates

China's bet on robotaxis didn't come out of nowhere. It's an extension of a deliberate, decade-long infrastructure build. As Socialblize notes, China's self-driving car firms have been given a headstart by the country's EV supply chain as they expand globally. That's not incidental—it's the whole strategic logic. When you've already industrialized the production of batteries, sensors, and the electric drivetrains that underpin autonomous vehicles, the marginal cost of adding the autonomy stack is lower than it would be for a competitor starting from scratch.

Companies like Baidu and Pony.ai are at the center of this push, operating driverless services across multiple Chinese cities and now positioning for international markets. The EV manufacturing base gives them a cost floor that Western autonomous vehicle startups—many of which are still burning through venture capital without a path to hardware scale—simply can't match on current trajectories.

The industry's own argument for why this matters globally runs something like this: if robotaxi operators can bring ride costs down to parity with—or below—a human-driven Uber, the addressable market expands dramatically, including for people with disabilities or mobility limitations who currently have limited transport options. It's a genuine argument. Whether the cost savings flow to consumers or to platform shareholders is a separate question that the industry tends to elide.

What "global expansion" actually means on the ground

Here's where the technology story and the labor story converge, and where the framing of "China's ambitious expansion" starts to feel incomplete.

Consider what a robotaxi rollout looks like from the other side of the windshield. I've been trying to reach Didi drivers in cities where the company has announced or expanded autonomous pilots. The specific interviews I'm still pursuing, but the structural situation isn't complicated to describe: gig drivers in China, like their counterparts in the United States, are typically classified as independent contractors. That classification means no unemployment insurance, no severance, no retraining obligation from the platform, and no union to negotiate a transition. When the work disappears, it disappears clean.

The global rideshare industry employs millions of people in this precarious category—Uber alone has somewhere between 3 and 4 million drivers worldwide—and that's before you count the traditional taxi workforce that gig platforms already disrupted once. A second disruption, this time automated, layers on top of a workforce that was already stripped of most labor protections during the first round.

The industry's cost-parity argument—making rides as cheap as or cheaper than a human-driven trip—is precisely what makes the displacement math so stark. Lower per-ride costs at scale means the economics favor removing the driver. That's the point. The people for whom cheaper rides expand mobility access are not the same people who currently depend on driving for income, and the robotaxi business model doesn't have an obvious mechanism for bridging that gap.

The data question nobody's asking cleanly

When Chinese robotaxi platforms eventually operate in Western cities—mapping streets, logging trip patterns, building behavioral profiles of urban movement—there's a question buried in the geopolitics that has a specific labor and civil liberties dimension: who gets surveilled, and who decides?

Concerns about Huawei's exclusion from 5G infrastructure in several Western markets, documented by Oxford Economics, established the regulatory precedent that data infrastructure with Chinese ownership warrants heightened scrutiny. Autonomous vehicles are, among other things, rolling data collection systems. Every trip logs location, passenger behavior, traffic patterns, and urban infrastructure detail.

The workers in this story—gig drivers who might be employed by or alongside these platforms during any transition period, or dispatchers who manage hybrid fleets—are also the people whose movements, labor patterns, and workplace conditions would be logged and processed by systems they have no visibility into and no negotiating power over. That's not a hypothetical concern about national security in the abstract. It's a concrete workplace surveillance question. American labor law has almost nothing to say about algorithmic management by foreign-owned platforms, and what little exists hasn't been tested against this model.

The EV playbook has limits

The optimistic version of the China-robotaxi story draws a straight line from EV dominance to robotaxi dominance: China scaled hardware manufacturing, used cost advantages to flood global markets, and will now do the same with autonomous mobility services.

The line isn't quite that straight. EVs are products—you manufacture them, export them, sell them, and the regulatory friction is mostly about import tariffs and safety certifications. Robotaxis are services, operating on public roads, subject to municipal licensing, insurance liability frameworks, and public trust in ways that a parked vehicle never is. The regulatory environments in the United States and Europe are fragmented, slow-moving, and in some places explicitly hostile to rapid autonomous deployment. A city council in Phoenix is not the same problem as a city council in Paris, and neither is the same problem as London's transport authority.

None of that makes the Chinese firms' position weak. It makes it more complicated than the EV analogy suggests, and the complications tend to fall hardest on workers and communities rather than on the companies doing the expanding.

The transition nobody's managing

What's missing from nearly every piece of coverage on autonomous vehicles—and what I keep returning to—is any serious account of what "the transition" actually means for the people it moves through.

Transitions have a way of being described, in business coverage and policy papers, as things that happen to markets. Robotaxis will "reshape urban transport." Autonomous vehicles will "redefine mobility." The passive constructions do a lot of work, absorbing the human cost into a clean narrative of progress.

What the transition actually means: a Didi driver in Chengdu who's been on the platform for six years, who hasn't been classified as an employee and therefore has no severance claim, watching the number of available rides contract as autonomous vehicles absorb the routes. A taxi dispatcher in a mid-sized American city whose job doesn't disappear in a single announcement but erodes shift by shift as the fleet mix changes. A gig worker who took on car payments specifically to qualify for higher-tier platform status, now holding a depreciating asset in a market that's moving against them.

The technology works—or is working well enough that the commercial deployment question is no longer hypothetical. The supply chain advantages are real. The cost trajectories are pointed in the direction the industry says they're pointed. What isn't built, anywhere in this picture, is any infrastructure for the people the transition will leave behind. No retraining pipeline, no portable benefits system, no organized worker power at the table where deployment timelines get set.

The robotaxi is coming. The question worth sitting with is who gets to decide how fast—and who absorbs the cost when the answer is: faster than workers can adapt.

---

Carmen Rodriguez covers labor, workplace organizing, and worker rights for Buzzrag.