New York Gig Worker Safety: 2026 AI Liability

It was a Tuesday afternoon in Manhattan when a delivery route went completely sideways for Mateo Rodriguez, a 32-year-old from Ecuador making a living as an UberEats cyclist in NYC. He was in the middle of a routine West Village run when the app’s AI suddenly pushed a new route on him, sending him down a street he knew was a death trap during rush hour. The resulting crash wasn’t just bad luck. It exposed the messy legal questions about who’s liable when a dispatch algorithm gets a human being hurt. It’s a serious question now: can you hold a piece of code accountable?

The Algorithm’s Imperative: A Delivery Goes Wrong

After almost two years cycling for UberEats, Mateo knew the city. He had his own routes, the ones that were actually safe, not just fast. He lived by his phone, which on this day sent him to pick up an order on Bleecker Street. The app first told him to take Christopher Street, a route he knew well. But right as he was making the turn, his screen flashed with a new order from the algorithm: “Optimal route updated. Turn right onto West 4th Street.”

Everyone who rides in the city knows that stretch of West 4th between 7th Avenue and 6th Avenue is a disaster, choked with double-parked trucks and aggressive cabs. His gut screamed no. But the app kept pinging, and ignoring it could ding his rating and cost him future jobs, so he listened to the machine instead of his experience. He made the right turn. A few moments later, he had to swerve around a van parked in the bike lane and a taxi making a sharp lane change clipped him, sending him and his bike skidding across the asphalt. His arm was twisted under him, and the pain was instant.

The Aftermath and the Gig Worker’s Bind

Paramedics from Lenox Health Greenwich Village got him stabilized and over to NewYork-Presbyterian Hospital. The damage was serious: a fractured ulna, bad road rash, and a concussion. His first thought, after the pain, was about money. As an independent contractor, he gets no paid sick leave and no workers’ comp, so being unable to ride means his income just stops. That’s the reality for these guys.

“I see cases like Mateo’s all the time,” says Sarah Chen, a New York City personal injury attorney who focuses on the gig economy. “These companies call their workers independent contractors to get out of employer responsibilities, but then they control every minute of their day with dispatch algorithms, performance ratings, and how they get paid. That control is what really muddies the water when someone gets hurt and we have to figure out who’s accountable.”

Working through New York’s Gig Economy Law

New York has made some moves to regulate this stuff. For example, New York City’s Local Law 115 of 2021 kicked in back in January 2022, setting up a minimum pay rate for food delivery workers that will eventually hit $17.96 an hour by April 2026 (before tips). That helps with pay, but it doesn’t change their independent contractor status or create any kind of injury compensation fund, which is the real problem here.

So when an UberEats cyclist gets hit in NYC, the main option is a standard personal injury law claim. For Mateo, that meant going after the taxi driver who was obviously negligent. But the AI dispatch adds a whole other layer. The bigger question became whether UberEats itself could be held partly responsible for sending him into that dangerous situation in the first place.

AI Dispatch and Negligence

This is where we get into the weeds of negligent algorithm design. To pin any liability on UberEats, Mateo’s lawyers had to show that the AI, in sending him down West 4th Street, created an unreasonable and foreseeable risk of harm. The point isn’t that the AI is ‘evil’. It’s that the company that built and deployed it failed in its basic duty of care.

“You have to start asking some hard questions,” Chen says. “Was the algorithm set up to chase speed no matter the safety cost? Does it even look at real-time traffic, accident data, or know which streets are magnets for crashes? If a platform’s AI keeps sending people down known high-accident corridors when there are safer options, you’ve got a strong argument for negligence. It all comes down to what a reasonable company would have done.”

Proving it is tough. You need experts who can testify about AI, analyze the data, and understand traffic safety patterns, which means you have to get your hands on UberEats’ internal data about how their algorithm actually works, something they fight tooth and nail to protect. One of the first things Mateo’s lawyers did was fire off a preservation letter to UberEats, telling them not to delete any data about his dispatch: the original route, the update, and whatever data the AI used to make its decision.

Gathering the Digital Breadcrumbs

Even in a lot of pain, Mateo had the presence of mind to screenshot the new route on his phone before the paramedics arrived, and that single action became a foundation of his case. From there, his legal team built out the file with everything they could find:

• Medical Records: Every note, scan, and report on his injuries and treatment from NewYork-Presbyterian Hospital.
• Police Report: The official write-up of the accident, with witness info and the initial take on the taxi driver’s fault.
• Eyewitness Accounts: Statements from people who actually saw the crash happen.
• Traffic Data: Reports showing historical accident rates and congestion on West 4th Street, proving it was a known problem area.
• UberEats App Data: All his trip histories, earnings statements, and any messages between him and the platform.

“You have to document everything,” Chen stresses. “Photos of the scene, your bike, your injuries, and especially screenshots of the app. That digital trail is often the only hard proof there is.”

The Legal Battle: A Two-Front Fight

The legal strategy for Mateo’s case had to attack on two fronts. The first part was simple: a standard personal injury claim against the taxi driver and his insurance for the unsafe lane change which is a pretty clear-cut case under New York’s Vehicle and Traffic Law. The second, much trickier part was the claim against UberEats, arguing their AI dispatch contributed to the crash by sending him somewhere they should have known was dangerous, which meant they breached a duty of care to him.

That whole argument against UberEats boiled down to one idea: even though Mateo was an independent contractor, the company still has a responsibility to make sure the tools it gives workers, like its dispatch algorithm, don’t put them in foreseeable danger. That responsibility comes from the control UberEats has over how the work gets done, regardless of what they call their workers.

This part of the law is still being figured out, and courts everywhere are trying to see how old tort principles fit these new scenarios with AI and algorithmic bosses. Luckily, New York’s CPLR Article 14-A (Comparative Negligence) allows a jury to split the blame between different parties, so even if the taxi driver was 90% at fault, a court could still find UberEats was 10% responsible and make them pay their share.

Resolution and Takeaways

Months of back-and-forth negotiations finally led to a settlement. The taxi’s insurance paid up, since their driver’s fault was obvious. The bigger deal was that UberEats also agreed to a confidential settlement, likely because they didn’t want to risk a long, public trial that could set a dangerous precedent for how their AI dispatch system is judged. It wasn’t an admission of guilt, but it shows these companies are feeling the legal heat over the safety of their tech.

Mateo’s recovery was a long haul of physical therapy and counseling. He did get back on his bike eventually, but he was a lot more cautious and had a much better understanding of his rights. His case offers some clear lessons for other gig workers:

1. Classification Isn’t Destiny: Being an “independent contractor” closes the door on some benefits, but it absolutely does not stop you from filing a personal injury claim against a negligent driver or even the platform itself.
2. Document Everything: Screenshot every app notification, route change, and message. Photograph every injury. Keep every record.
3. Get a Lawyer Immediately: You need an attorney who gets the gig economy to protect the evidence and find every possible defendant. Don’t wait.
4. The Algorithm Is a Target: As AI directs more of our work, its role in causing accidents will get more and more attention from the courts. It’s a real factor.

This whole incident with the UberEats cyclist in NYC is a clear signal that our laws have to catch up to the technology. Chasing efficiency with AI is one thing, but it can’t come at the expense of the safety of the people who actually have to do the work.

Conclusion

If you’re a gig worker hurt on the job in New York City, your path to getting fair compensation depends on you knowing your rights and documenting every single detail, especially the instructions the app’s algorithm gave you.