Why Federal Regulators Just Let a Robotaxi Ditch the Steering Wheel
Four Seats, No Wheel, One Signature
Chloe Huu stepped out of a boxy white vehicle on the Las Vegas Strip Thursday afternoon and tried to describe what she’d just felt. "It’s kind of scary at first, but when I just got used to it, it’s really cool and chilling," she said, mentioning the adjustable temperature and the music almost as an afterthought, the way you’d praise legroom on a flight you weren’t sure would land. What she’d actually just done was ride in a car with no steering wheel, no driver’s seat in the traditional sense, and four seats facing inward toward each other like a small lounge on wheels — a vehicle that, under a fresh piece of robotaxi regulation handed down that same day, was legally allowed to charge her money for the privilege.
That distinction matters more than it sounds. Amazon’s Zoox unit had been giving these rides away for free in Las Vegas and San Francisco since the previous year, a kind of extended public trial. Thursday’s approval from federal regulators changed the transaction itself: Zoox can now sell rides in vehicles that were never built with a human driver in mind, pending state and local sign-off. People were already lined up at stops along the Strip waiting for one, even with a projected wait time that kept sliding around in ways Huu found genuinely annoying.
The Rulebook Nobody Wrote for This
The reason a robotaxi regulation announcement out of Las Vegas made national news has less to do with one company’s product launch than with a much older problem: American vehicle safety law was never built for a car that drives itself. Every federal safety standard governing steering columns, brake pedals, mirrors and wipers assumes, as a starting premise, that a person sits behind the wheel. Jonathan Morrison, administrator of the National Highway Traffic Safety Administration, put it bluntly. "If you’re going to be developing a standard in the 1970s, 1980s, 1990s, nobody at that time had this imagination, nobody had a crystal ball that you might have technologies that would enable the vehicle to drive itself," he said.
That’s the short version of why this is surfacing now: NHTSA proposed a rule last month that would let robotaxis operate without brake pedals altogether, a proposal still pending as regulators try to catch decades-old rules up to a technology that has already left the lab. Zoox’s Thursday approval to charge for rides, and the earlier clearance to remove steering wheels, are pieces of that same catch-up effort — not one company’s victory lap, but a live rewrite of what a legal car is even required to have.
What a Car Doesn’t Need When Nobody’s Driving It
Once you accept that premise, the argument for removing a steering wheel stops sounding radical and starts sounding almost obvious. Morrison’s own framing was that things like steering wheels, windshield wipers and side-view mirrors simply aren’t needed in a vehicle that drives itself — they’re vestiges of a design language built around human hands and human eyes, not around cameras, lidar and machine perception.
But regulators aren’t simply taking a company’s word for that. NHTSA’s approval process is meant to function, in Morrison’s words, as the "cop on the beat to make sure that this is being done right" — a role that means Zoox has to earn trust in stages rather than all at once. The company is only permitted to deploy up to 2,500 vehicles without steering wheels over the next two years, a cap explicitly designed to let the technology prove itself before it scales. Amazon’s ambitions run considerably larger: the company hopes to eventually produce as many as 10,000 robotaxis a year at a plant near Silicon Valley, a manufacturing bet that only makes sense if the regulatory ceiling keeps rising alongside it.
More than half a million people have already ridden in Zoox vehicles, according to the company, which is the kind of number that sounds reassuring until you ask what, exactly, was being measured. That question turns out to be the crux of the entire regulatory fight.
The Data Nobody Can Compare
Peter Kurdock, general counsel for the Advocates for Highway and Auto Safety group, raised a pointed objection to Zoox’s steering-wheel-free plan: the company, in his view, hadn’t provided detailed evidence that its vehicles could operate safely on public roads without those controls. "Lots of promises, lots of glossy language, lots of ‘We believe, we think we’ll do this.’ Yet there really was no data behind the promises, and that’s really concerning to us," Kurdock said.
The Insurance Institute for Highway Safety has flagged a structural version of the same problem. The limited crash data the government currently requires autonomous vehicle companies to submit makes it genuinely hard to assess how safe these vehicles actually are as they expand into more cities. Most companies don’t report how many miles their vehicles have driven, which means there’s no denominator — and without a denominator, there’s no crash rate, just a raw count of incidents that means little on its own.
There’s a second layer buried in that gap that’s easy to miss: the reporting standard itself is inconsistent. There’s no clear rule for which incidents autonomous vehicle companies must disclose, so these companies end up reporting far more minor incidents than human drivers do — not necessarily because robotaxis crash more, but because a human driver typically only reports a crash when there’s real damage or injury, while an autonomous vehicle operator may log a low-speed bump or a sudden stop that a person would never have called in at all. The two data sets aren’t measuring the same thing, which means the public conversation about "how safe is a robotaxi" is often comparing numbers that were never built to be compared.
Weird Isn’t the Same as Dangerous — Except When It Is
Morrison doesn’t dispute that robotaxis have had problems. Some are almost comic in isolation — a vehicle stopping in the middle of the road for no visible reason, or struggling to find its way out of a parking lot, the kind of glitch that reads as software awkwardness rather than danger. Others are not comic at all. NHTSA has opened investigations, pushed recalls and issued calls to action in response to incidents of autonomous vehicles passing stopped school buses or interfering with first responders — the exact scenarios where a split-second human judgment call is supposed to happen and, in an automated system, has to happen inside code instead.
"Sometimes we have seen unsafe behaviors, and we jump on that really quickly again. That’s our job at NHTSA," Morrison said. He was careful, though, to frame the technology’s promise alongside its failures rather than let the failures define it. "By and large we’re really excited about the promise of this technology and we want to see that promise fulfilled," he said, pointing to a specific upside that’s easy to state and hard to argue with: a robotaxi never drives distracted, and it never drives drunk, two of the most persistent causes of fatal crashes among human drivers.
The Company Trying to Catch a Ten-Year Head Start
It’s worth sitting with why Zoox, and not some other name, is at the center of this particular approval. Amazon paid $1.2 billion to acquire the self-driving startup, a bet that only pays off if Zoox can close a real gap: Waymo, owned by Google’s parent company Alphabet, already operates in multiple cities and has years of operational lead time that Zoox and Tesla are both still working to close. In an industry where the actual product is trust — trust that a car with no wheel and no pedals will do the right thing at the right moment — a head start compounds. Every mile driven without incident builds the case for the next approval; every incident resets some of that case to zero.
Zoox CEO Aicha Evans framed Thursday’s approval in exactly those terms. "This achievement reflects a shared commitment to enabling innovation while maintaining the highest standards of safety," she said — a sentence that’s really doing two jobs at once, claiming momentum for the company while also, deliberately, keeping NHTSA’s safety framing inside the same breath. That’s not incidental phrasing. It’s the language of a company that knows its next approval depends on the last one holding up under scrutiny.
Why the Regulation Race Is the Real Business Story
This is where robotaxi regulation stops being a legal footnote and becomes the actual business of the industry. The vehicle itself — the sensors, the routing software, the inward-facing seats — is table stakes. What separates a company that can put 2,500 cars on the road from one still waiting on approval is the regulatory relationship it has managed to build, and that relationship is now the primary competitive asset in the entire sector. A company that can demonstrate a clean, well-documented safety record earns looser caps and faster scaling; a company that leans on promotional language without data, the exact criticism Kurdock leveled, risks the opposite outcome.
Seen this way, the current patchwork of state and local approval requirements sitting on top of the federal one isn’t bureaucratic friction so much as a second, slower-moving test that every robotaxi operator has to pass twice. Zoox’s federal green light lets it charge for rides; it still needs the state and local layer before a single paying customer gets in. That two-tier structure means the fastest path to scale isn’t necessarily the best engineering — it’s the best evidence trail, built one uneventful ride at a time across half a million trips and counting.
What Riders Are Signing Up For
For an ordinary rider, none of this regulatory architecture is visible from the back seat, which is precisely the point of the system working correctly. What a rider notices instead is closer to what Huu described: novelty, comfort details like climate and music, and a practical annoyance — a wait-time estimate that shifts around enough to make planning difficult. That’s a small complaint, but it points to something real about where this technology currently sits: the safety case is being built methodically and slowly, while the actual user experience is still catching up to the reliability people expect from an app-summoned ride.
The practical takeaway for anyone watching this space, whether as a future rider or simply as someone trying to understand where transportation is heading, is that the interesting signal isn’t the friendly headline about a fun ride on the Strip. It’s the reporting gap the IIHS identified — the absence of a shared, comparable safety standard across an entire industry that’s expanding into more American cities every year. Until autonomous vehicle companies report miles driven and incidents on the same terms, the public will be asked to trust a safety record that can’t actually be measured against the human drivers it’s meant to replace.
What Happens the Next Time Something Goes Wrong
The honest ending to this story isn’t a verdict, because there isn’t one yet. NHTSA has built itself a role as both enabler and referee — approving the removal of steering wheels while reserving the right to recall, investigate, and cap deployment numbers the moment something goes wrong. That balance holds only as long as regulators keep matching each new capability with proportionate evidence, and only as long as companies like Zoox keep treating half a million incident-free rides as a floor to build on, not a finish line.
The steering wheel Zoox removed was never really the point. The point is whether the industry can produce the kind of data that lets a stranger get into a car with no driver and reasonably believe the numbers behind it. Right now, that number doesn’t fully exist. It’s being written, ride by ride, on the Las Vegas Strip.
Vantrue N4S 3 Channel Dash Cam
This product may be useful for you: Vantrue N4S review: triple-channel 2.7K dash cam with STARVIS 2, PlatePix GPS and IR night vision — full buyer evidence
As an Amazon Associate, TruePickUS can earn from qualifying purchases.
FAQ
What is robotaxi regulation and why is it changing now?
Robotaxi regulation refers to the federal, state and local rules governing how autonomous vehicles can operate, including whether they need steering wheels or brake pedals. It’s changing because most existing vehicle safety standards were written decades ago under the assumption that a human would always be driving, and regulators like NHTSA are now updating those rules to match vehicles that drive themselves.
Can robotaxis legally operate without a steering wheel?
Under a recent federal approval, Zoox’s autonomous vehicles can operate without a steering wheel, though the company is currently capped at deploying up to 2,500 such vehicles over two years while it demonstrates safety. A separate proposed rule would extend similar treatment to brake pedals, but that proposal remains pending.
Is it safe to ride in a robotaxi?
Regulators and safety groups differ somewhat on this. NHTSA has said it investigates unsafe behaviors quickly and remains largely optimistic about the technology’s potential, while groups like the IIHS and Advocates for Highway and Auto Safety note that inconsistent crash-reporting standards across companies make it difficult to fully assess robotaxi safety records compared with human drivers.
Why don’t autonomous vehicle companies report a clear crash rate?
Most companies don’t disclose how many total miles their vehicles have driven, which is needed to calculate a meaningful crash rate. There’s also no standardized rule for which incidents must be reported, so companies tend to report more minor incidents than human drivers typically would, making direct comparisons unreliable.
How does Zoox compare to Waymo in the robotaxi market?
Waymo, owned by Alphabet, already operates in multiple cities and holds a significant head start in both regulatory approvals and operational experience. Zoox, backed by Amazon’s $1.2 billion acquisition, is working to close that gap, with plans to eventually manufacture up to 10,000 robotaxis a year.
Related Reading
- Amgen’s Data Breach Shows Why Your Hospital’s Cloud Vendor Matters More Than Its Firewall
- NAVAPAL Car Wash Brush: Reach, Squeegee & Soft Bristles
Official Source Links
For further reading and verification: