Massachusetts rideshare drivers now have something drivers in most of the country do not: an actual certified union sitting across the table from Uber and Lyft.
That makes the App Drivers Union important. It also means the ADU deserves scrutiny.
The question should not simply be whether the union gets drivers another raise. The real question is much larger:
What hills is the App Drivers Union willing to die on?
Because the first contract negotiated in Massachusetts could become a model that gets copied elsewhere. If the union gets this right, Massachusetts could establish meaningful limits on how rideshare platforms control drivers. If it gets it wrong, Uber and Lyft could end up with a unionized system in which they still retain most of the power that caused drivers to organize in the first place.
Massachusetts certified the App Drivers Union, SEIU 32BJ/IAM, as the exclusive bargaining representative for the state’s transportation-network drivers on May 22, 2026. That gives ADU the authority to bargain over wages, benefits and other working conditions for an industry-wide bargaining unit. (Massachusetts Government)
Actual negotiations with Uber and Lyft began August 12.
ADU says drivers have organized around better pay, fair deactivation procedures, transparency around algorithms, safer working conditions, dignity and “keeping human drivers behind the wheel.” (App Drivers Union)
Those are good objectives.
Now comes the harder part.
A Union Has to Do More Than Negotiate a Number
Massachusetts drivers already have an earnings floor resulting from the Attorney General’s settlement with Uber and Lyft.
So if the major accomplishment of the first ADU contract eventually becomes, “We negotiated another couple dollars,” I am going to have questions.
Pay matters. Of course it matters.
But the fundamental problem with app-based work is that compensation is only one piece of the platform’s control.
Uber can determine what jobs are shown to a driver.
It can determine what information drivers see before accepting them.
It can design incentives around acceptance.
It can change classifications.
It can send drivers into different types of work.
It can determine how cancellations are handled.
It can decide what information remains visible after a trip.
And increasingly, these companies can decide whether a human driver is involved at all.
A serious rideshare union has to bargain over those systems.
Otherwise, you can negotiate the greatest hourly guarantee imaginable and watch the company redesign the work around it.
Courier Work Is One of Those Tests
This is one area where I think Massachusetts should pay very close attention.
There is a major difference between delivering somebody’s hamburger around the corner and putting somebody’s property in your vehicle and taking it three cities away.
That second job is courier transportation.
If I accept a package and spend an hour driving it across Massachusetts, my vehicle has effectively been occupied by that customer’s property for an hour.
I cannot put a rideshare passenger in that seat at the same time.
I may also finish that delivery well outside the market where I intended to work.
That creates lost opportunity, additional mileage and the possibility of unpaid repositioning back to where passenger demand exists.
Economically, that starts looking a hell of a lot more like transporting a passenger than dropping off a cheeseburger.
Yet Massachusetts’ current union framework contains an important limitation: ADU represents transportation-network drivers while they are performing covered rideshare work. The state specifically says the union does not represent food-delivery work such as Uber Eats. (Massachusetts Government)
Fine.
But then somebody has to define where actual courier transportation fits.
Because the platform should not be able to avoid rideshare protections merely by replacing the human being in my back seat with somebody’s shit.
If Uber wants exclusive use of my automobile for a transportation assignment lasting 45 minutes, 60 minutes or 90 minutes, that should be treated differently from a five-minute restaurant delivery.
That is exactly the kind of classification issue a drivers’ union should be fighting over.
The App Itself Has to Be Bargained Over
Another hill should be affirmative consent.
If I turn on UberX, that does not necessarily mean I volunteered to become a food-delivery driver, package courier, pet transportation provider or anything else the company decides might increase utilization.
Turning down a popup should mean no.
Closing a popup should not mean yes.
A driver should be able to explicitly select which services the driver is willing to perform, and those choices should remain selected until the driver changes them.
That sounds like a small technical issue.
It isn’t.
The app is the workplace.
For a factory worker, a union might negotiate what happens on the production floor. For a rideshare driver, a large part of the production floor is software.
That means algorithmic transparency cannot merely mean receiving an explanation after something goes wrong.
Drivers should be able to see what happened.
What category was this offer?
What was the original estimated mileage?
What destination was displayed?
What was the estimated compensation?
When was it accepted?
When was it canceled?
How much active time was credited?
Did declining the previous offer affect what was offered next?
Was a service category enabled, and when was it enabled?
Those records should not disappear just because the trip is over.
If a company is calculating somebody’s livelihood through software, the worker needs an audit trail.
Deactivation Has to Mean Due Process
ADU has already identified fair deactivation procedures as one of its priorities. (App Drivers Union)
That should be another major hill.
A driver should not lose access to income because an automated system received an accusation and decided something happened.
There should be notice.
There should be evidence.
There should be an opportunity to respond.
And there should be an actual human review before somebody loses their livelihood for any significant period.
That becomes even more important when platforms increasingly rely upon automated fraud detection, ratings, telematics and algorithmic enforcement.
A union grievance system that merely helps somebody navigate the same opaque system Uber already controls would not be enough.
The point should be introducing an independent check on platform power.
Then There Are Autonomous Vehicles
This is where I will give ADU credit.
They are talking about it.
ADU publicly says one of its goals is keeping human drivers behind the wheel. (App Drivers Union)
Good.
Now I want to know how hard that position becomes when something has to be traded at the bargaining table.
Because this may ultimately be the biggest hill of all.
There isn’t much point organizing 70,000 drivers, negotiating their wages, protecting their deactivation rights and improving their working conditions if the companies’ long-term answer is:
That’s great. We’re replacing you.
Autonomous vehicles are not simply another technological improvement for this particular workforce.
They are potential labor substitution.
ADU therefore has every reason to make autonomous deployment a labor issue.
That doesn’t necessarily mean Massachusetts can or should prohibit every autonomous vehicle from ever operating on a public roadway.
But commercial passenger transportation is different.
A state can make decisions about whether companies should be permitted to replace large portions of an existing commercial transportation workforce with autonomous fleets, particularly while questions involving safety, liability, regulation and public-road use are still being resolved.
And a union representing rideshare drivers should unquestionably be part of that argument.
The important question will be whether “keeping human drivers behind the wheel” remains a slogan or becomes an enforceable principle.
Not Every Issue Can Be the Hill
Any negotiation requires compromise.
ADU isn’t going to get everything drivers want.
Uber and Lyft aren’t going to get everything they want either.
That is bargaining.
But that makes the choice of non-negotiables extraordinarily important.
Maybe the union accepts a smaller immediate pay increase in exchange for real algorithmic transparency.
Maybe it fights harder for cancellation compensation.
Maybe it makes affirmative consent for different work categories mandatory.
Maybe courier transportation becomes its own protected classification.
Maybe there is an enforceable deactivation procedure.
Maybe autonomous substitution becomes the fight around which everything else revolves.
Those choices tell drivers what kind of union ADU intends to become.
And Massachusetts law puts additional pressure on this process. If bargaining has not produced an agreement within 180 days of ADU’s May 22 certification—November 18, 2026—either side can request mediation. If mediation subsequently fails, Massachusetts law provides a path toward arbitration, but drivers can ultimately face a vote involving arbitration or decertification. (Massachusetts Government)
So this isn’t an endless theoretical discussion.
Decisions are going to have to be made.
The Rest of the Country Should Be Watching
Massachusetts is becoming one of the laboratories for the future of rideshare labor.
That means drivers outside Massachusetts should be paying attention too.
Not because Massachusetts necessarily has all the answers.
Because whatever happens here gives everyone else evidence.
We are going to learn what Uber and Lyft concede when drivers collectively bargain.
We are going to learn what they refuse to concede.
We are going to see whether algorithmic management can actually be negotiated.
We are going to see whether the industry attempts to move work between classifications.
We are going to see whether long-distance courier work gets treated as real transportation work.
And eventually we are going to find out how seriously organized rideshare labor intends to confront autonomous replacement.
The first ADU contract therefore shouldn’t be judged merely by whatever dollar figure ends up on the front page.
Read the rest of it.
Look at what happens to deactivations.
Look at classifications.
Look at courier work.
Look at cancellation compensation.
Look at driver consent.
Look at algorithmic transparency.
Look at autonomous vehicles.
Most importantly, look at what ADU was willing to walk away from the table over.
Because every union can produce a list of things it would like to have.
You find out what an organization actually represents when you discover which hills it is willing to die on.