TEAMSTERS VS DRIVERLESS

SO WHAT HAPPENS TO THE DRIVER?

DRIVERLESS TRUCKS ARE HERE.

THE CONVERSATION TRUCKING NEEDS TO HAVE—NOT SOMEDAY. NOW.

A Follow-Up to ADPI-005: Leadership Isn’t About Being Liked—It’s About Getting Results

Take the driver out of an 80,000-pound truck and the first question most people ask is whether the computer can drive it.

I think trucking has been asking the easier question.

The harder one is what happens to the person whose skill, paycheck and identity used to occupy that seat.

That is no longer a question for someday. Commercial freight is already moving without a human controlling the truck from the cab in limited operations. The technology has crossed the line from demonstration to business model, and the empty seat is part of that model.

Sean O’Brien’s answer is to keep a trained human inside the truck. I understand why. My answer is different: I do not believe America can afford to stop developing autonomous trucking. Somewhere between those positions is the conversation this industry keeps postponing.

In ADPI-005, I judged O’Brien by one standard: WHAT CHANGED FOR THE PEOPLE HE REPRESENTS BECAUSE HE LED? His strongest answer was not a speech, a political appearance or a viral moment. It was enforceable language—something a worker can point to after the cameras leave.

Now that leadership model is meeting a technology capable of changing the occupation itself.

THE EMPTY SEAT IS NOW A BUSINESS MODEL

Drivers have heard predictions about autonomous trucks for years. Some were breathless. Some were ridiculous. Deadlines passed, companies disappeared and plenty of people concluded that the whole thing was another Silicon Valley sales pitch.

That conclusion is no longer safe.

Aurora launched commercial driverless freight between Dallas and Houston in April 2025. By the end of June 2026, the company said its trucks had traveled nearly 440,000 driverless miles, and it was operating across 10 routes in the Sun Belt. These were not conventional driver-assist trucks with CDL drivers waiting to take over. There was no person behind the wheel. Those are company-reported results, but they are real commercial operations—not a concept video. (Aurora’s 2025 launch; July 2026 update)

Kodiak is proving the model in a different environment. At the end of the second quarter of 2026, it reported 35 customer-owned driverless trucks, more than 40,000 hours of paid driverless operation and more than 20,000 cumulative loads. Most of that work has been in Atlas Energy’s industrial network in the Permian Basin, not unrestricted coast-to-coast freight on public interstates. Kodiak says it is targeting a public-road long-haul driverless launch by the end of 2026. That distinction matters. (Kodiak AI)

PlusAI is aiming at factory-built autonomous trucks and a 2027 commercial launch. In August, it reported what it calls a Safety Case Readiness score of 93.4 percent and continued work with International and Ryder along the I-35 corridor in Texas. That number is the company’s own readiness measurement—not a government finding that the truck is “93.4 percent safe.” But it shows how quickly this is moving from experimental equipment toward trucks intended to be built, sold and placed in fleet service. (PlusAI)

None of this means a driverless semi can go anywhere a professional driver can go. It cannot enter every customer, back into every dock, read every confused hand signal at an emergency scene, cross every mountain in every storm or improvise its way through every piece of bad information American trucking can throw at it.

These systems operate inside defined conditions—mapped routes, approved roads, weather limits and operating rules. But one technical question has been answered: a Class 8 truck can move commercial freight on public highways without a human controlling it from the cab UNDER SPECIFIC CONDITIONS.

The question now is how far those conditions can expand, how safely they can expand and whether they can expand CHEAPLY ENOUGH to change the economics of trucking.

That last part will drive this faster than the novelty ever could.

A human driver needs sleep, compensation and time away from work. Some receive health insurance and retirement benefits. We get sick. We have families. We reach Hours-of-Service limits because fatigue is real and public safety requires boundaries.

Software does not sleep or reach the end of a human Hours-of-Service clock. A driverless truck still needs maintenance, sensors, mapping, communications, remote support, cybersecurity, insurance and people throughout the operation. Machines fail too. They simply fail in different ways.

The business bet is that all of those costs will eventually be less than the compensation attached to the driver’s seat—and that the equipment can earn money during more hours of the day. If that bet proves correct, pressure to scale will not come only from engineers fascinated by the technology. It will come from carriers and shippers studying the numbers.

WHAT SEAN O’BRIEN HAS ACTUALLY DONE

O’Brien is not waiting for the economic argument to settle itself.

The Teamsters’ federal autonomous-vehicle policy says a human operator should remain in every autonomous vehicle regardless of its automation level. It also calls for federal safety standards, cybersecurity requirements, crash reporting, workforce-impact studies, wage replacement and real retraining protections. The union’s position rests on two concerns that cannot be separated: JOBS AND SAFETY. (International Brotherhood of Teamsters)

The most important part of O’Brien’s record is that the Teamsters have carried that position beyond policy papers.

At TForce Freight, the five-year agreement running through July 2028 prohibits the company from using robots, driverless vehicles, drones or other technology to move freight or replace covered drivers, clerks and dockworkers. (TForce Freight agreement)

THAT IS CONTRACTUAL PROTECTION.

It does not stop autonomous technology across America, and it does not protect every driver. It protects the workers covered by that agreement because they have a representative with the legal standing and bargaining power to make the company sign its name.

If I were a driver wondering whether software might replace the job feeding my family, I would know the difference between a politician saying he stands with me and a contract saying what the employer cannot do.

That is where O’Brien is strongest.

It is also where my position begins to separate from his. I support using human operators as a safety and employment bridge while the technology and the evidence are still developing. I do not know that keeping a person inside every autonomous truck forever is the permanent answer. If a system eventually proves it can operate more safely than a qualified professional driver within a defined environment, a permanent human mandate could become a job-preservation rule disconnected from the safety evidence.

But anyone who wants to remove that mandate should have to prove the case with more than a press release.

SAFETY CANNOT BE SETTLED BY THE COMPANY GRADING ITS OWN EXAM

Autonomous developers have a legitimate argument. Sensors do not drink, text, become impatient or fall asleep. Software does not decide to push another 50 miles after its eyes begin closing. Professional drivers know better than anyone how much danger comes from fatigue, distraction, impairment and reckless human judgment.

Removing those risks does not remove risk itself.

What happens when cameras are blinded by glare or road spray? When a sensor is damaged or obstructed? When lane markings disappear in construction? When a police officer gives directions that conflict with the map? When communications fail, software behaves unexpectedly or the truck becomes disabled in a live lane? Who inspects the equipment before departure? Who secures the scene after a failure? How quickly can remote assistance respond? What prevents a cyberattack from becoming a highway emergency? And who carries legal responsibility when the machine makes the wrong decision?

The industry has safety cases, simulations, testing programs and growing operating records. All of that is evidence worth examining. It is not yet a final verdict.

The commercial Class 8 driverless record is still tiny compared with the billions of miles accumulated by human-driven trucks across every kind of road, traffic, weather and mechanical condition. The comparison is also difficult because companies operate in different environments and publish different measurements.

On July 30, 2026, NHTSA announced a three-year effort to accelerate development of the nation’s first-ever autonomous-vehicle performance standards. That does not prove driverless trucks are unsafe. It does show that the national yardstick for judging them is still being built while commercial deployment is moving forward. (NHTSA)

The public should not have to choose between corporate confidence and labor fear. We need transparent, comparable and independently examined evidence: miles, crashes, disengagements, roadside failures, operational limits, emergency response and what happens when the system encounters something it was not designed to handle.

A machine should not be required to achieve perfection when human drivers do not. It should be required to show that it can perform at least as safely as a qualified professional driver within every condition where a company wants permission to remove that driver.

That is not anti-technology. It is the price of putting 80,000 pounds on a public road without a human being in control.

CALIFORNIA IS WHERE THE ARGUMENT BECAME REAL

California shows what happens when technology, labor and government collide.

The Legislature passed a bill in 2023 that would have required a trained human operator in autonomous vehicles weighing more than 10,000 pounds. Gov. Gavin Newsom vetoed it, arguing that the state’s regulatory process could address safety. Labor kept fighting while regulators continued building a path toward heavy-duty autonomous operations.

In April 2026, the California DMV adopted regulations allowing manufacturers to apply for permits to test and eventually deploy heavy-duty autonomous vehicles. The process is phased. It begins with testing with a safety driver, then moves toward driverless testing and commercial deployment. The regulations require 500,000 miles in each testing phase, a structured safety case, expanded reporting and emergency-response capabilities. They did not place unlimited driverless semis on every California highway the next morning. They created the legal road for companies to get there. (California DMV)

Then, on August 5, Teamsters California sued the DMV.

The lawsuit alleges that the agency failed to conduct the economic review required by state law and did not adequately examine the rules’ effects on jobs and safety. The union wants the regulations invalidated. The Autonomous Vehicle Industry Association says regulators spent years taking public input, and the DMV has declined to comment while the case is pending. Those are the parties’ positions; no court has decided the claims. (Teamsters; CBS Sacramento)

Whatever happens in court, the lawsuit tells us something important about O’Brien and the Teamsters. They are not merely speaking against automation. They have bargained contract language, backed legislation, challenged regulators and now supported a legal attempt to force an economic accounting before the trucks are deployed.

They will not win every fight. They may not be able to stop this technology. But they are forcing the worker into a room that might otherwise be occupied entirely by developers, carriers, manufacturers, investors and government officials.

Look around that room.

Autonomous companies have engineers. Mega carriers have executives, consultants and attorneys. Shippers have procurement departments. Wall Street has analysts. Manufacturers have research divisions. Government has regulators.

WHO REPRESENTS THE DRIVER?

That is why O’Brien matters even when I disagree with him.

“THERE WILL BE OTHER JOBS” IS NOT GOOD ENOUGH

Whenever automation threatens an occupation, somebody eventually says workers can retrain. Technology creates other industries. People will adapt. There will be new jobs maintaining sensors, monitoring fleets and assisting trucks remotely.

Some of that will be true. It is also incomplete.

If one remote operator can assist several trucks, one new job does not replace several driving jobs. If the new work is located hundreds of miles away, requires different credentials or pays less, telling a displaced driver that the economy created another position somewhere does not make him whole.

If a company removes a driver because software can perform the work more cheaply, the plain fact remains:

HIS JOB DISAPPEARED.

There is another loss that does not fit easily into a workforce-impact study. Some people actually LOVE DRIVING TRUCKS.

Some of us love the machinery. We love the road and the independence. We love learning how a truck handles and knowing when something does not feel right. We love putting a trailer into a difficult dock and getting it right. We love knowing that most people could not climb into that seat tomorrow and perform the job safely and professionally.

For many drivers, trucking is not simply an occupation. IT IS A CRAFT. After enough years, it becomes part of who you are. Telling that person, “We will retrain you,” does not fully answer what may be taken away.

Think about somebody who loves an old American muscle car. You can put him in an electric vehicle that accelerates faster, requires less routine maintenance and carries far more computing power. By almost every measurable number, the new vehicle may outperform the old one.

Then ask him whether the experience is the same.

It is not.

The sound is different. The machinery is different. The connection is different. He is not rejecting the stopwatch. He is telling you the stopwatch did not measure everything that mattered.

Work can be the same way. A paycheck matters, but work can also provide pride, identity, skill, independence and accomplishment.

Sometimes EFFICIENCY DOES NOT REPLACE MEANING.

That is the part of this argument the balance sheet will never volunteer.

WE CANNOT UNINVENT THE TRUCK—SO NEGOTIATE BEFORE THE LAYOFF NOTICE

I do not support a national strategy that begins and ends with BAN THE TRUCKS. Once a transformative technology exists, America cannot make it disappear by refusing to develop it while the rest of the world keeps moving.

Autonomous trucking is connected to artificial intelligence, robotics, sensors, semiconductors, computer vision, logistics, advanced manufacturing and defense. The capabilities will move between industries whether trucking welcomes them or not. Walking away would not protect America from the technology. It would influence who develops it, who controls it and who gains from it.

But “you cannot stop progress” is not a labor policy.

If autonomous trucking keeps expanding—if the systems become safer, more capable and less expensive—then the industry should negotiate the transition BEFORE drivers receive the layoff notice.

That could mean autonomous operation on proven interstate corridors while humans continue handling complex urban environments, customer interaction, final-mile work and conditions the system cannot safely manage. It could mean requiring a trained operator until transparent data show that the system performs safely without one in that specific operation—not everywhere in theory, but there in fact.

It should mean giving experienced drivers first opportunity at work involving inspections, remote assistance, safety review, training, maintenance and autonomous-fleet operations. Those opportunities must be tied to real positions with disclosed pay, location and qualifications—not a brochure telling a 58-year-old driver to learn computers.

It should mean advance notice, income and benefit protection, company-funded training and bargaining over how new technology changes the work. Where jobs are eliminated, the companies capturing the savings should carry a meaningful share of the transition cost.

And it should mean telling the truth: new technology jobs may not replace driving jobs one for one. A transition plan is not credible if its math depends on pretending every displaced driver will become a remote operator or robotics technician.

I do not pretend to have the entire solution. But I know FOR DRIVERLESS TRUCKS OR AGAINST DRIVERLESS TRUCKS is the wrong question.

The real questions are where the technology has proved itself, who verifies that proof, who is accountable when it fails and what is owed to the people whose work creates the savings.

DO NOT STOP THE TECHNOLOGY—BUT DO NOT ABANDON THE DRIVER

Sean O’Brien is willing to push much harder toward keeping a human physically inside every autonomous commercial truck than I am. I understand why. His job is to represent workers, and his record shows that he is willing to turn that responsibility into contracts, legislation and legal action.

My position is this: America should keep building.

We should lead in artificial intelligence, robotics, autonomous transportation, logistics and advanced manufacturing. We should require honest safety evidence and permit autonomous systems where that evidence shows they can perform the work. We should compete and win.

But winning a technology race should not require treating American workers as disposable.

THOSE TWO IDEAS SHOULD NOT BE ENEMIES.

Professional drivers cannot protect their future by hoping autonomous trucking fails. Some companies will fail. Some predictions will prove wrong. Some work may remain human-operated for generations. But hope is not a plan.

Neither is telling millions of working people to get out of the way.

Progress is never free. Somebody pays for it. Sometimes the company. Sometimes the consumer. Sometimes the taxpayer. And sometimes it is the person who spent 25 or 30 years mastering a trade only to learn that the industry has decided the trade no longer needs him.

That should bother us—not enough to stop America from advancing, but enough to plan for the people carrying the cost of that advancement.

Perhaps O’Brien and I are standing at different places on the same road. He is saying, DON’T REMOVE THE DRIVER. I am saying, DON’T STOP THE TECHNOLOGY—BUT DON’T ABANDON THE DRIVER EITHER.

Those are different positions. Underneath them is one shared principle:

THE HUMAN BEING BEHIND THE WHEEL MATTERS.

I do not want America to lose the race in autonomous trucking. I also do not want the professional driver treated as roadkill on the way to winning it.

We can build. We can innovate. We can lead. And we can remember the people who carried this economy before a computer learned how to drive the truck.

Because once the driver is removed, saying there are plenty of other jobs does not answer the hardest question:

WHAT DID WE DO WITH THE PERSON WHO USED TO BE BEHIND THE WHEEL?

That is the conversation trucking needs to have.

NOT SOMEDAY. NOW.

— Will Cook

A Driver’s Perspective