The Take — Uber's war on robotaxis is a toll booth, not a conscience

Uber is not trying to stop robotaxis. It is trying to make sure that when they arrive, they arrive on Uber's platform and pay Uber a cut. Every piece of the company's new labor politics — the union alliances, the workforce-transition language, the 85 percent rule its lobbyists are pushing in New Jersey — is aimed at that outcome, not at slowing the technology down.

I think the distinction matters more than it sounds, because it changes who we should root for. If this were a company making common cause with workers against a machine that replaces them, the labor movement would have found an unlikely ally. What's actually happening is a middleman using worker protection as the argument for a rule that would hand it the robotaxi market.

Read the rule, not the press release

The tell is in the mechanism. In New Jersey, Uber's lobbyists have pushed a requirement that human drivers handle 85 percent of rides on any platform offering robotaxi service. That is not a brake on autonomy. It is a quota that forces Waymo and Tesla onto a network that already supplies human drivers at scale — Uber's. The company that spent $59 million to keep drivers classified as contractors is now arguing that drivers need protecting, and the protection it wants happens to route all autonomous supply through its own dispatch layer.

The timing strips away the rest of the cover. On Wednesday Uber announced it is cutting 3,300 jobs, roughly 10 percent of staff and its largest reduction since 2020, with CEO Dara Khosrowshahi framing it as flattening management layers to speed up decisions. The same company plans to put more than $10 billion into robotaxis. A firm that genuinely believed autonomous vehicles were coming too fast would be shrinking its bet on them, not doubling it while shrinking its headcount to pay for it.

The workforce argument isn't fabricated, and that's what makes it effective. Uber's director of AV and AI policy told the D.C. council that the legislation there "largely ignores the workforce transition," and pointed out that one AV in California now performs the work of roughly four drivers. That is a real displacement rate and a real gap in the law. It is also an argument whose proposed remedy — cap the autonomous share — leaves Uber as the toll collector on every trip, robot or human.

The counter-case, stated fairly

The strongest version of Uber's position starts from the map rather than the motive. The company's CFO says it is reacting to bills that would have produced no AVs at all, and he has a point: the alternative to a bad deployment bill in a state legislature is often no deployment at all, and a blanket ban helps nobody except incumbent taxi interests. On that reading, Uber is negotiating the least-bad version of a transition that is happening regardless, and the 85 percent figure is a bargaining position rather than an end state.

There is also a genuine public-interest case for a quota. A workforce transition that throws hundreds of thousands of drivers out of work with no sequencing is a real harm, and nobody else with Uber's leverage is making it. Unions taking Uber's money and lobbying help are not naive — they can read the rule too, and they have decided that 85 percent of rides going to humans is better than the alternative their members would otherwise get. If the outcome is more human rides than the market would have delivered on its own, the drivers win whether or not Uber's motives are clean.

And the competitive picture is genuinely uncertain. Waymo is in 14 cities with more than 4,000 vehicles and over 500,000 paid rides a week, but it is fighting a tariff clock on the Chinese-built vans that carry its newest expansion, as our analysis of Waymo's Ojai rollout laid out. Tesla's purpose-built Cybercab is still mostly an event. Nobody can say with confidence who operates the fleets in 2030, which is exactly when a platform-agnostic rule looks prudent rather than cynical.

Why the take still holds

Motive tests are usually a waste of time in policy fights, and I'd skip this one if the rule were neutral about who captures the value. It isn't. A rule that caps autonomous share per platform does not slow autonomy; it taxes it, and the tax is collected by the company with the largest human-driver supply. That is the definition of a toll booth, and the fact that some drivers also benefit from it doesn't change what the structure is built to do.

The sequencing is what convinces me. Uber announced the layoffs and the robotaxi spending in the same breath as the union alliance, which means the company has already decided where the industry is going. What it has not decided is whether it gets paid on the way. As our morning brief on the union alliance and the job cuts reported, the same week produced a 10 percent staff reduction and a $10 billion autonomous commitment — a company that believes AVs are coming is not trying to stop them, it is trying to own the curb.

There's a second reason to be skeptical of the solidarity framing: Uber has spent the last two years making itself the most efficient operator in the room precisely by cutting the cost of everything it touches. Its own engineering org reportedly grew AI agent use 9.4-fold since February while holding total AI spending flat since April, which we covered in Uber's AI agent use is up 9.4x while spending flatlines. That is a company with real discipline about unit economics, and it is the same instinct on display in Trenton.

The unions are the ones I'd watch. They are taking a partner of convenience in a fight where their interests and Uber's genuinely overlap right now. That overlap ends the moment autonomy is cheap enough that Uber would rather run its own fleet than broker someone else's — and at that point the 85 percent rule becomes a cost rather than a shield.

What would change my mind

Two things, both observable. First, if Uber starts backing rules that slow autonomy outright — fleet caps, per-city limits, geofences that shrink supply — rather than rules that shape who supplies it. That would be a genuine anti-AV position, and it would cost Uber money, which is the only reliable test of sincerity. Second, if the company puts its own robotaxi capital behind the human-driver quota in a way that visibly sacrifices its autonomous roadmap. It hasn't: the $10 billion is still there.

I'd also revise if the 85 percent figure turns out to be unions' number rather than Uber's, drafted by the people whose work is at stake and merely carried by the company with the lobbyists. Who authored the language is checkable, and it would change the story. Until someone produces that draft, assume the entity paying for the campaign wrote the rule.

When a platform funds a labor protection that routes the work through that platform, is that solidarity or capture? Tell us in the comments.

Sources: Financial Times — Uber allies with driver unions in bid to slow robotaxi rollout · CNBC — Within Uber-Waymo split, a key labor battle over AVs is being waged in the nation's capital · USA Today — Why Uber is cutting 3,300 jobs as robotaxi competition grows · Axios — Uber's AI usage jumps 9.4x while spending stays flat · Techmeme — Uber is aligning with drivers' unions