On September 3, 2026, Tesla did something historically important.
It began commercial deployment of Cybercab in Austin, Texas - a purpose-built autonomous vehicle without a permanently attached steering wheel, accelerator pedal, brake pedal or conventional side mirrors.
The same day the National Highway Traffic Safety Administration opened an investigation into Tesla's certification of the vehicle.
"NHTSA investigates Tesla" makes for an ominous headline.
But read what NHTSA itself has been saying about autonomous vehicles, and a much more interesting story emerges.
Because just six months before Cybercab's launch, NHTSA Administrator Jonathan Morrison publicly described almost the exact regulatory problem Cybercab is now forcing the government to confront.
His basic point was extraordinary: Vehicles of the future are being constrained by regulations written for vehicles of the past.
And NHTSA knows it.
That means AQ26002 could ultimately become something very different from a roadblock to Cybercab.
It could become the process through which the United States finally establishes how a vehicle designed never to have a human driver fits into a regulatory system built entirely around one.
If Tesla's interpretation survives that scrutiny, Cybercab could emerge with something considerably more valuable than a temporary exemption: regulatory precedent.
What Actually Happened
Tesla informed NHTSA that it had certified Cybercab as complying with all applicable Federal Motor Vehicle Safety Standards, or FMVSS.
The important word is applicable.
Cybercab does not have permanently attached conventional controls such as a steering wheel, brake pedal, accelerator pedal or conventional mirrors. Nevertheless, Tesla determined that the vehicle could be self-certified under the existing federal framework.
On September 3, NHTSA opened Audit Query AQ26002 to examine the process and technical data Tesla used to reach that conclusion.
NHTSA subsequently explained that the inquiry will examine whether Tesla's compliance framework relies on determinations that certain FMVSS requirements are simply inapplicable to Cybercab.
We know that NHTSA...
- Has not determined that Cybercab is noncompliant.
- It has not ruled that Cybercab must have a steering wheel.
- It has not ruled that Cybercab must have pedals.
- It has not said Tesla must obtain an exemption.
- And it has not ordered Tesla to halt Cybercab deployment.
America Does Not Pre-Approve Every New Car
One reason the story is easily misunderstood is that many people assume NHTSA "approves" a new automobile before it reaches consumers. That generally isn't how the American system works.
Manufacturers are responsible for certifying that their vehicles comply with applicable Federal Motor Vehicle Safety Standards.
NHTSA establishes and enforces those standards and retains the authority to investigate whether a manufacturer's certification was valid.
NHTSA explained the system itself in announcing the Cybercab inquiry: automakers certify that their vehicles satisfy federal requirements, subject to subsequent agency oversight.
So Tesla did not submit Cybercab to NHTSA and wait for the government to approve it.
Tesla certified Cybercab. NHTSA is now auditing that certification.
NHTSA Predicted This Exact Fight
This is where the story becomes fascinating.
On March 10, 2026, NHTSA Administrator Jonathan Morrison addressed the National AV Safety Forum in Washington.
NHTSA Administrator Jonathan Morrison - https://www.nhtsa.gov/nhtsa-leadership/jonathan-morrison
He explained that America's FMVSS were intended to be technology-neutral - but were inevitably written around the technology that existed when they were created.
And for essentially the entire history of federal vehicle regulation, every automobile had one thing in common: a human driver.
Consequently, the standards assume that human being exists.
Morrison gave braking as an example. Existing light-duty braking standards are framed around stopping performance after a specified force is applied to a physical foot pedal.
Then he described the central problem in remarkably direct language: "The technology of the future is often tethered to the regulations of the past."
That could almost be the subtitle of the Cybercab investigation.
Does a Driverless Car Become Safer Because It Has a Pedal?
Morrison then made an observation that goes directly to the heart of Cybercab. A vehicle designed never to be operated by a human, he said, "doesn't get safer because it has a foot pedal."
Think about the significance of that statement. The head of NHTSA was openly acknowledging that requiring a physical human control on a vehicle that will never have a human driver can become a regulatory artifact rather than a genuine safety requirement.
That doesn't mean NHTSA has already decided Tesla's Cybercab complies with existing law. It hasn't. But it does mean that NHTSA understands the underlying problem Tesla is confronting.
The issue is not necessarily: Should autonomous vehicles be allowed to avoid safety requirements?
The issue is: Which requirements actually produce safety benefits when the human driver around whom those requirements were constructed no longer exists?
The Steering Wheel Is Not the Safety System Anymore
Consider how radical this transition really is.
For more than a century, the automobile's safety architecture assumed that a human controlled the machine.
- The steering wheel translated human intent into lateral motion.
- The accelerator translated human intent into propulsion.
- The brake pedal translated human intent into deceleration.
- Mirrors provided information to the human so that the human could make decisions.
Cybercab changes the architecture.
- Software determines direction.
- Computers make driving decisions.
- Electronic systems command acceleration and braking.
- Sensors perceive the environment.
The autonomous driving system effectively replaces the human driving loop. That does not eliminate the need for safety regulation.
Quite the opposite.
It means safety standards eventually need to regulate the things that actually make an autonomous vehicle safe rather than mechanically preserving interfaces intended for a driver who no longer exists.
NHTSA itself recognizes this transition.
The Regulatory Paradox
Morrison called the situation a paradox.
NHTSA understands that some legacy requirements may not create additional safety in a purpose-built autonomous vehicle. Yet those requirements remain legally binding until they are changed, interpreted as inapplicable, or an exemption is granted.
That is precisely the tension AQ26002 now has to resolve.
Tesla's argument is not: "We violate the safety standards, but please let us operate anyway."
Tesla's position is that Cybercab complies with all standards applicable to Cybercab, while some requirements written specifically around conventional human controls do not apply to its architecture. NHTSA is now testing whether Tesla is right.
Cybercab self-certification sticker
And Then NHTSA Said Something Even More Interesting
Morrison's next observation may be even more relevant to Tesla.
NHTSA does have authority to grant exemptions for new technology.
But Morrison warned: "Reliance upon the generosity of a regulator is a poor business model."
Read that sentence again in the context of Cybercab.
It describes precisely why obtaining temporary exemptions vehicle by vehicle is unlikely to be the ideal foundation for a transportation platform Tesla hopes eventually to manufacture by the millions.
An exemption can be useful. It can allow experimentation. It can enable early deployment. But a company intending to create an enormous autonomous transportation network probably does not want its long-term business model dependent upon repeatedly receiving discretionary regulatory permission to build its fundamental product.
Tesla appears to be pursuing a much more consequential path.
Rather than saying: Cybercab doesn't comply, therefore exempt us.
Tesla is saying: Cybercab complies with the federal standards that legally apply to Cybercab.
If Tesla can successfully defend that position, the implications are enormous.
This Is Why the Investigation Could Actually Be Good
At first glance, federal scrutiny of a brand-new vehicle sounds undesirable. But consider the alternative. Imagine Tesla deployed 10,000 Cybercabs. Then 100,000. Then one million. And throughout that expansion, the fundamental legality of Tesla's FMVSS interpretation remained unresolved.
Every rollout would carry the same question. Every state. Every city. Every lawsuit. Every serious accident. Opponents could repeatedly argue that Cybercab's underlying federal certification remained legally vulnerable.
Instead, NHTSA is forcing the issue to be confronted at the beginning of the ramp.
There were only a small number of Cybercabs in commercial deployment when NHTSA opened AQ26002. The agency's opening document acknowledges that Tesla intends gradually to add vehicles and locations.
If there is a fundamental problem with Tesla's interpretation, discovering it early is far better than discovering it after millions of vehicles have been manufactured.
But the reverse is also true.
If Tesla is right, resolving the issue now could remove one of the largest regulatory uncertainties hanging over Cybercab before the vehicle reaches massive scale.
NHTSA Isn't Trying to Stop Autonomous Vehicles
The broader remarks from the National AV Safety Forum make another point clear.
NHTSA's stated policy is
not to prevent
autonomous vehicles from reaching the market.
Morrison described the agency's strategy as one of "deliberative urgency" - advancing technology while maintaining rigorous safety oversight.
NHTSA's federal automated-vehicle framework has three stated pillars:
- Prioritize the safety of AV operations on public roads.
- Remove unnecessary regulatory barriers to safe innovation.
- Enable commercial AV deployment to improve safety and mobility.
NHTSA can simultaneously investigate Tesla aggressively
and
want Cybercab-like technology to succeed.
Those positions are not contradictory.
The agency's job is to establish where the safety boundaries should actually lie.
NHTSA Is Already Rewriting the Rulebook
NHTSA says it is reviewing the FMVSS for unintended barriers created by vehicle designs enabled by automated driving systems.
At the March forum, Morrison specifically discussed regulatory changes involving transmission shifting, windshield defrosting and windshield wipers, with additional proposals in development.
By September, when NHTSA announced the Cybercab investigation, the agency said eight rulemakings were underway, including standards involving brake pedals, windshield wipers, lighting and rearview mirrors.
Think about that. Two of the most controversial missing components on Cybercab - pedals and mirrors - are among the areas NHTSA already says require regulatory modernization.
NHTSA also explicitly says that it wants to remove unnecessary barriers to American AV innovation. The complication is that until those new rules are finalized, today's requirements remain in force. And that is exactly why Tesla's interpretation matters so much.
Cybercab Is Forcing Tomorrow's Question Today
For years regulators could discuss purpose-built autonomous vehicles as a future problem. Now one is operating commercially in Austin. Cybercab turns a theoretical debate into an immediate regulatory question.
Instead of asking: How should we eventually regulate a vehicle without a human driver?
NHTSA now has to ask: How do our existing regulations apply to one operating today?
That pressure can accelerate regulatory evolution. Sometimes regulation changes first and technology follows. Other times technology arrives first and forces regulation to catch up.
Cybercab may be a textbook example of the latter.
The Zoox Precedent Shows the Risk Is Real
None of this means Tesla is guaranteed to win. Zoox provides an important warning.
Zoox also developed a purpose-built autonomous vehicle without conventional driving controls and self-certified that vehicle as compliant with applicable FMVSS. NHTSA opened Audit Query AQ23001 in March 2023 to investigate the basis for that certification. NHTSA inspected Zoox vehicles and reported multiple apparent FMVSS noncompliances. Zoox subsequently pursued an exemption, and NHTSA closed AQ23001 in August 2025 after that exemption was granted.
NHTSA even described Zoox as an industry outlier because it had originally self-certified its purpose-built vehicle rather than seeking an available exemption.
That precedent demonstrates that an Audit Query is real regulatory scrutiny - not ceremonial paperwork. Tesla will have to prove its case. But Tesla also has something Zoox did not have when its original dispute began: several additional years of regulatory development and a NHTSA leadership publicly acknowledging that legacy human-control requirements can create inappropriate barriers for vehicles designed never to be human-driven.
Tesla Could Establish the Path Everyone Else Follows
This is where the potential upside becomes enormous.
Suppose NHTSA examines Cybercab's engineering. It examines Tesla's interpretation of each relevant FMVSS. It reviews Tesla's technical evidence. It challenges Tesla's reasoning. And ultimately Tesla's certification survives.
Cybercab would then have accomplished something far more important than simply avoiding an exemption. It could establish a credible regulatory pathway for an entirely new class of automobile.
The automobile industry spent more than a century building machines around the human driver.
Tesla is effectively making the argument that: the vehicle itself is now the driver.
If federal regulations can accommodate that architecture without special exemptions, that is a major historical transition.
It Could Give Cybercab Greater Legitimacy
Autonomous vehicles require people to surrender something humans have controlled since the birth of the automobile: the act of driving. That demands trust.
Tesla saying Cybercab is safe is one thing. Tesla's certification surviving intensive federal scrutiny would be another.
If NHTSA examines Tesla's engineering, certification methodology and technical assumptions and ultimately concludes that Tesla's approach complies with federal law, the investigation itself becomes part of Cybercab's credibility.
It Could Destroy One of the Biggest Cybercab Bear Arguments
There has been a persistent argument surrounding Cybercab:
- Tesla can't really sell it without a steering wheel.
- Federal law requires pedals.
- The test vehicles with controls prove Tesla will eventually put them back.
- Tesla will need special exemptions.
- Therefore the purpose-built Cybercab cannot scale in its intended form.
AQ26002 has the potential to settle that question far more convincingly than another Tesla presentation ever could.
If NHTSA rejects Tesla's interpretation, Tesla may have to modify Cybercab, obtain exemptions, alter its certification strategy or wait for regulations to change. That would be significant. It should not be dismissed. But consider the other outcome.
If NHTSA accepts Tesla's interpretation that Cybercab legally complies with every FMVSS applicable to its design?
One of the largest regulatory bear cases surrounding Cybercab could collapse. And it wouldn't collapse because Tesla said so.
It would collapse after federal scrutiny.
Why Not Simply Ask for an Exemption?
This is perhaps the most important question. Zoox ultimately obtained one. Why wouldn't Tesla simply do the same?
NHTSA Administrator Jonathan Morrison has effectively provided one possible answer himself: A business designed around continual regulatory indulgence is not an ideal business model.
A temporary exemption may be perfectly appropriate for limited demonstrations or transitional deployments. It is a very different proposition to base a potentially multi-million-unit manufacturing platform on one.
Tesla's ambitions for Cybercab are enormous. If Cybercab eventually becomes a mass-produced autonomous transportation platform, Tesla presumably wants a durable regulatory foundation beneath it.
That makes the difference between compliance and exemption enormously important.
An exemption says: The rules apply, but we're allowing you not to comply with some of them under specified conditions.
Tesla's apparent position says something much more powerful: Those particular human-control requirements don't apply to this vehicle in the first place.
AQ26002 may determine whether Tesla can successfully make that argument.
NHTSA Wants Regulatory Certainty Too
Toward the end of his March remarks, Morrison discussed the need to align federal, state and local approaches as autonomous vehicles move from limited pilots toward multi-state commercial deployment.
The goal, he said, is a shared framework that preserves safety and public trust while providing the regulatory certainty industry needs to scale nationwide.
That sentence may ultimately prove extraordinarily relevant to Cybercab. Because regulatory certainty is precisely what Tesla needs if Cybercab is going to evolve from a handful of vehicles in Austin into a national transportation network.
AQ26002 could contribute to creating that certainty. Even if the process is uncomfortable.
The Outcome Matters More Than the Investigation
There will undoubtedly be headlines:
- NHTSA Investigates Tesla.
- Federal Government Questions Cybercab.
- Tesla Robotaxi Faces Regulatory Scrutiny.
Those headlines are technically true. But they don't tell you the outcome. NHTSA has not concluded Tesla is wrong or that the Cybercab has been declared noncompliant. Tesla has not been ordered to install a steering wheel, nor ordered to install pedals. And Tesla has not been told to halt deployment.
This Could Be the Investigation Tesla Wants to Win
Tesla should not want weak regulation. It should want clear regulation.
Weak regulation leaves safety questions unresolved. Unclear regulation leaves business models vulnerable. Clear regulation tells companies exactly what they must achieve and allows capital, engineering and manufacturing to move with confidence.
The Cybercab investigation could help move America toward that clarity. If Tesla loses the argument, we will learn that Cybercab's regulatory architecture needs to change.
But if Tesla wins?
Then AQ26002 could become one of the most important regulatory events in Tesla's history.
It could validate the idea that a purpose-built autonomous vehicle does not need to mimic the physical architecture of a human-driven automobile merely because that is how automobiles have always been built. It could remove an enormous uncertainty surrounding Cybercab. It could create a regulatory moat. It could establish precedent for an entirely new class of vehicle. And it could help accelerate the rewriting of federal vehicle standards for the autonomous age.
Six months before Cybercab arrived, America's top vehicle-safety regulator said that the technology of tomorrow was being tethered to regulations from yesterday.
Cybercab is forcing America to write the rules for the future.
