ANALYSIS: LiDAR Mandatory in New Jersey; Tesla’s Camera Strategy Is Wavering

Two sensors are better than one

In contrast to Tesla, other players have taken the opposite approach, and New Jersey seems to be proving them right. Waymo, a subsidiary of Alphabet, and Zoox, owned by Amazon, are piling on sensors like one would stack guardrails. Cameras, LiDAR, radar: each technology compensates for the other’s blind spots. When the camera is blinded by low-angle sunlight, the LiDAR continues to measure. When rain blurs the lens, the radar cuts through the curtain. This redundancy is expensive; it adds weight to the vehicle and complicates the engineering. And yet, it offers something that minimalism can never promise: a second chance when the first fails. When it comes to driverless driving, that second chance is not a luxury. It’s the difference between a reported incident and an irreparable tragedy.

New Jersey isn’t just copying this model. It’s making it the standard. By requiring “cameras and at least two other types of detection technologies,” the state is turning an engineering preference into a legal requirement. This is a choice with far-reaching consequences. It means that a legislator—not the market—decides the architecture of an autonomous vehicle. Some will cry foul over bureaucratic interference. Others will see it as the only rational response to an industry that tests its hypotheses on public roads, amidst citizens who haven’t given their consent. Between corporate freedom and the protection of pedestrians, the state has made its choice. And it has sided with the person crossing the street without knowing they’re part of an experiment.

Redundancy is not a sign of an engineer’s cowardice. It is the memory of all the accidents we didn’t want to count.

Each state is moving forward on its own

The most mind-boggling aspect of this situation is what’s missing. The United States still lacks a comprehensive federal framework for autonomous vehicles. In 2026, the world’s leading technological power is letting each state cobble together its own regulations. California requires permits and public reporting. Texas is taking a much more flexible approach. New Jersey could become the most interventionist of all, dictating part of the onboard equipment. The result: a robotaxi that’s perfectly legal in Austin could be illegal in Newark. The same car, the same technology, the same risk—and fifty different responses. This fragmentation isn’t just an administrative detail. It’s an admission of political impotence.

Who benefits from this lack of a common framework? Manufacturers who know how to navigate between jurisdictions, choosing permissive states and avoiding strict ones. It leaves citizens in complete the dark about what’s driving alongside them. And yet, America likes to present itself as the laboratory of the future, the place where innovation asks no one for permission. Autonomous driving reveals the flip side of this pride: when everyone goes it alone, no one protects the whole. Europe, often mocked for its regulatory red tape, at least has the merit of a clear direction. In fact, global regulations were adopted in late June 2026. The United States, on the other hand, prefers a patchwork. A patchwork where every missing tile is a blind spot for public safety.

A country that allows fifty states to write fifty laws does not have fifty freedoms. It has a single void, repeated fifty times.

Challenging the Law Rather Than Changing the Hardware

Faced with this threat, Tesla has not backed down. The company is already lobbying against the New Jersey bill. Its argument is clever: the proposal does not focus on actual safety performance, but would impose a technical architecture that would exclude it from the market. Translation: don’t judge our sensors; judge our results. The argument deserves to be heard. A law that locks in a specific technology can indeed stifle innovation and punish the one who might be right against all odds. And yet, there is something troubling about seeing a company prefer to challenge a safety law rather than add a sensor. Modifying its hardware seems, in the very words of the analysis, “highly unlikely.” Stubbornness takes precedence over adaptation.

This is where the debate becomes political. Who should write the rules for a technology that operates among living people? The engineer who designs it, or the legislator who protects those who encounter it? Tesla would like measured performance to replace the hardware requirement. On paper, the idea is appealing. In reality, however, it requires blind trust in data produced by the company itself. Yet industrial history is rife with firms that swore their figures proved their innocence—until a report proved otherwise. New Jersey doesn’t trust that. It demands physical, tangible safeguards that are independent of the manufacturer’s promises. Given the choice between a company’s word and a sensor that sees in the dark, lawmakers have chosen the sensor.

A company that would rather rewrite the law than add a sensor tells us what it’s really protecting. And it’s not us.

1,000 taxis announced, 42 on the road

There are the promises, and then there are the numbers. Tesla aimed for a fleet of 1,000 robotaxis. It appears to have deployed only 42. Let that contrast sink in. Nine hundred fifty-eight vehicles missing between the promise and the road. This isn’t just a delay. It’s a chasm between the narrative and reality. For years, Elon Musk has been announcing full autonomy “for next year” with a regularity that has come to resemble a literary genre. And yet, the fleet remains minuscule, and troubling figures have emerged regarding the behavior of these robotaxis. Autonomous driving isn’t a matter of faith. It’s a matter of kilometers driven without killing anyone. On this front, the promise stands exposed.

These 42 vehicles reveal a truth that marketing obscures. Deploying a safe robotaxi is infinitely harder than selling it at a conference. Every challenging situation, every ambiguous intersection, every unpredictable pedestrian tests the system’s limits. It is precisely here that the camera alone reveals its flaws, and where New Jersey asserts its standards. We can admire Tesla’s boldness, its ability to shake up a complacent industry. We might even hope it’s right, because an affordable robotaxi would make autonomous mobility accessible to everyone. But admiration doesn’t make a car drive. These 42 vehicles, however, are on the road. And their pitiful number—compared to the 1,000 that were announced—reminds us that the promised future hasn’t arrived yet. It’s still waiting. Maybe just one more sensor.

A thousand promises, forty-two cars. The rest is just hot air being sold at the price of the future.

From Attention Monitoring to Identity Verification

While the battle over external sensors rages on, another shift is brewing inside the cabin. A piece of code spotted in the latest update to Tesla’s iOS app reveals a feature currently in development: verifying the driver’s identity using the in-cabin camera before allowing the Full Self-Driving software to be activated. Until now, this camera has been used primarily to monitor the driver’s attention. Identity verification would serve an entirely different purpose. The shift is subtle but profound. We’re moving from a safety tool—ensuring the driver is watching the road—to an access control tool—verifying who you are. And this, inevitably, raises concerns that the company won’t be able to dispel with a simple press release.

Such a feature would make it possible to control who can activate a paid option: FSD subscription, fleet, rental vehicle, or a future robotaxi service. The business logic is crystal clear. But it reopens the question of hardware—one that haunts Tesla like a shadow. The in-cabin camera remains a conventional sensor. It is not an advanced biometric system with depth mapping, like the facial recognition on a high-end smartphone. In other words, Tesla wants to perform serious biometric authentication with a tool that wasn’t designed for that purpose. It’s the same gamble as with the exterior: asking a simple sensor to do a complex job. The “all-camera” strategy doesn’t stop at the windshield. It’s invading the cabin. And everywhere, it raises the same troubling question: How much can you cut corners on hardware before reliability takes a hit?

A camera that used to monitor your attention now wants to scan your face. The same eye, a new purpose. No one asked you if you agreed.

What an Extra Sensor Is Really Worth

Let’s get down to the nitty-gritty. On one hand, cost savings on hardware that make the robotaxi profitable. On the other, an extra sensor that sees in the dark, in the rain, and amid urban chaos. Tesla is prioritizing cost. New Jersey is focusing on safety. This isn’t a technical debate. It’s a moral trade-off disguised as an engineering specification. How much is the certainty worth that a car will see a child darting out between two parked vehicles at dusk, when the camera is struggling against the fading light? Tesla’s answer: AI will suffice. New Jersey responds: Add a LiDAR—you don’t mess around with this. And yet, both sides invoke safety. One through software performance, the other through hardware redundancy. The word is the same. The stakes are opposite.

What makes this debate so heated is that it isn’t playing out in a closed laboratory. It’s playing out on open roads, amid pedestrians, cyclists, and children who don’t know they’re part of a real-world test. When a company chooses to cut corners on a sensor, it shifts part of the risk onto the person crossing the street. This shift is invisible and painless—until an accident occurs. So the question isn’t whether Tesla is technically right or wrong. The question is who bears the burden of doubt while we search for the answer. Today, it’s not the shareholder. It’s not the engineer. It’s the stranger on the sidewalk. New Jersey has decided to give that person back a little protection. You could call that bureaucracy. You could also call it decency.

Cutting corners on a sensor simply shifts the risk. And the risk always falls on the person who didn’t sign anything.

Who Writes the Rules for the World to Come

This dispute in New Jersey goes far beyond the automotive industry. It raises a question that the West will have to confront everywhere: who writes the rules for autonomous vehicles? The company that manufactures them, driven by speed and profit? Or the democratic state, which is slower and more cautious but acts as the guardian of the common good? Silicon Valley insists that innovation doesn’t ask anyone’s permission. It’s a nice slogan for a T-shirt. It’s a dangerous doctrine for society. Because the very same players who reject U.S. regulation meekly bow to the demands of other markets when money dictates it. The freedom they proclaim is often a selective freedom. And yet, Western democracy has a strength that authoritarian regimes lack: it can say no to its own giants.

Look at China. There, the state and the tech giants march in lockstep, with no countervailing power, no public debate, and no ordinary citizen who can challenge them in court. Efficiency there is brutal, deployment is massive, and surveillance is total. Faced with this model, the West has a duty to prove that we can innovate without crushing citizens, and move forward without sacrificing caution. New Jersey, with its small local law, is part of this demonstration. It reminds us that an elected lawmaker can still require a billionaire to install a sensor. It’s a modest step. It’s essential. For the day our democracies give up on writing the rules for their own machines is the day they give up on something greater than a debate over LiDAR. They give up on deciding what kind of world they live in.

A democracy that no longer dares to say no to its tech giants has already begun to resemble the very entities it claims to fight.

What if stubbornness were the right approach?

Let’s be completely honest: there’s no proof that Tesla is wrong. The history of technology is littered with dogmas that seemed absurd before becoming common knowledge. Perhaps camera-based AI will one day reach a level where LiDAR seems like a costly and unnecessary redundancy—a relic of outdated caution. Perhaps Elon Musk—as detestable as his penchant for provocation may be—will have proven right against the entire industry. That would be the triumph of stubbornness. And we must give Tesla credit for one thing: without its radical approach, the electric car wouldn’t have taken off so quickly, and the world would still be running on oil for much longer. We can despise the man and still respect the upheaval he has caused.

But respecting boldness does not mean we should fail to monitor it. A gamble can be brilliant and still remain a gamble. Until Tesla has proven—with public and independent data to back it up—that its camera alone matches the redundancy of its competitors, doubt remains legitimate. And when lives are at stake, that doubt naturally leans toward caution. New Jersey isn’t telling Tesla, “You’re wrong.” It’s saying, “Prove it first, then protect.” This is the opposite of the typical Silicon Valley approach, which is to roll out first and prove later—at the expense of users. If Tesla wants to convince people, the path is clear: 80,000 kilometers without incident, transparent reports, and a public demonstration. The burden of proof lies with the innovator, not with the regulator. That’s how a self-respecting society works.

We may hope that Tesla is right. But we cannot assume it is right until it has proven it. That subtle distinction separates trust from naivety.

The One Whose Name No One Dares to Mention

Amid all this battle over sensors, lobbying, shareholders, and iOS codes, someone is missing. He has no lawyer, no press office, no seat on the board of directors. It’s the pedestrian—the one crossing an intersection in Newark on a rainy evening, unaware that a camera is struggling to make out his silhouette against the wet asphalt. No one mentions his name in press releases. And yet, he is the real issue at stake. This entire technical dispute boils down to a question that no communications professional asks: Will the car see him in time? LiDAR answers yes with physical certainty. The camera answers “maybe,” with statistical confidence. Between the “yes” and the “maybe,” there is a body crossing the street.

It is for this anonymous figure that New Jersey wrote its law. Not to annoy Tesla. Not to favor Waymo. So that the evening pedestrian does not become the adjustment variable in an economic model. We tend to treat autonomous driving as a topic for engineers and billionaires. That’s a mistake. It’s a topic of sidewalks, crosswalks, and children running after a ball. Every sensor removed from the vehicle is a safety margin taken away from someone who was never consulted. The debate seems abstract, full of numbers, and technical. In reality, it’s terribly concrete. It plays out at ankle height, right where a wheel stops—or doesn’t stop. That’s why a small law from New Jersey deserves to be discussed far beyond its borders.

We debate LiDAR and radar. We forget the child crossing the street. It’s always the same oversight, and it always comes at the same cost.

When One State Inspires Fifty

A local law is only local in appearance. If New Jersey adopts this requirement for sensor redundancy, it sets a precedent that other states could invoke. California, already strict, could tighten its rules. New York, Massachusetts, and Illinois could follow suit. And suddenly, the American mosaic would take on a new shape. The redundancy camp would gain legal ground, state by state, until the “all-camera” strategy becomes commercially unsustainable across a growing portion of the country. This is how industry dogmas die: not all at once, but through the gradual closing of doors. Tesla knows this. That’s why its lobbying efforts in New Jersey aren’t an isolated battle. It’s a fight to prevent the first domino from falling.

This mechanism of precedent is the quiet weapon of federal democracies. In the absence of a common law, each state becomes a testing ground for influence. What is enacted in Trenton can be replicated in Sacramento, then in Albany. And yet, this slowness has a downside: while states copy one another, the industry deploys, tests, and gains the upper hand. The law is always playing catch-up with technology. New Jersey is trying to reverse this race, to impose regulations before deployment rather than after an accident. It’s the opposite bet from Tesla’s—a bet on caution versus a bet on speed. We’ll know in three years, at the end of the pilot program, which of the two has saved the most lives. In the meantime, one thing is certain: the debate won’t stay confined to New Jersey. It’s already spreading elsewhere.

A well-written, concise law travels farther than a grand speech. It crosses state lines like an idea that can no longer be caught.

Technology is never neutral

At its core, the New Jersey case isn’t about sensors. It’s about values. It asks what we’re willing to entrust to machines—and at what cost. It asks who—the engineer or the citizen—has the final say. It asks whether progress must move forward at any cost, or pause long enough to ensure it isn’t trampling on anyone. These questions aren’t limited to self-driving cars. They concern artificial intelligence, facial recognition, and the automation of our entire lives. The car is just the tip of the iceberg. What we decide here—regarding the LiDAR on a robotaxi—foreshadows what we’ll decide tomorrow on far broader issues. Technology is never neutral. It always reflects the priorities of those who design it.

That is why this debate deserves more than just a spot in the auto section. It is a test of maturity for our democracies. Will they be able to enforce their rules on their own giants, without stifling innovation or failing to protect the vulnerable? New Jersey offers an imperfect, debatable, and perhaps overly rigid response. But it is a response. One state dared to say, “Here are our terms,” in the face of one of the world’s most powerful companies. At a time when so many leaders kowtow to tech billionaires, this simple act of courage deserves recognition. It reminds us of a truth we too often forget: democracy is not just the right to vote. It is the right to tell a giant that it, too, must install a sensor. Just like everyone else.

The day our democracies stop demanding one more sensor from their billionaires is the day they will cease to be democracies. The rule that protects the weak is the last bulwark against the reign of the strong.

Caution is not the enemy of the future

New Jersey hasn’t passed its law yet. Tesla hasn’t lost its market yet. Nothing is decided yet. But something has already shifted. One state dared to make the future of a giant contingent on the presence of LiDAR. It reminded us that safety isn’t an option you can cut to improve profit margins. And it has raised—perhaps unwittingly—the defining question of our time: How much control do we let machines have, and who remains in charge? Tesla embodies boldness, risk-taking, and faith in software. New Jersey embodies caution, safeguards, and the protection of the vulnerable. Both have a valid point. But when a child crosses the street in the dark, caution has a decisive advantage: it doesn’t gamble with a life.

So we can hope that Tesla proves its camera is enough. We can hope for a future of autonomous, clean, affordable mobility that’s accessible to all. But that future won’t be built on a promise repeated a thousand times for forty-two cars delivered. It will be built on evidence, on kilometers driven, and on rules written by elected officials—not by marketing departments. New Jersey has just driven this point home with a small law that carries great weight. One question remains, and I’ll leave it to you. On the day your car decides on its own whether or not to brake for a pedestrian crossing the street, would you rather it see with certainty, or guess with confidence? Your answer speaks volumes. About it. About us. About the world we’re willing to let roll on.

One more sensor means one less doubt. And when in doubt, it’s always the pedestrian who pays the price. Remember that the next time someone tries to sell you the future at a discount.

Signed, Maxime Marquette, columnist