The driver’s name was Michael Wallace. He was 37, a Kentucky native, engaged to be married, a father of four. He drove a brand-new Nissan Pathfinder in Cayenne Red, and he kept a handgun—a birthday present from his father—in the driver’s-side door pocket. In the course of more than 4,600 trips driving for Uber, he had never used it. And then, he did.
A little after 5:30 p.m. on Saturday, August 4, 2018, Uber routed Wallace to a Louisville Marriott to pick up a group of five. One of them was Jon Strand. Strand was in town from Indianapolis as part of his work at a company specializing in window installation. His wife, Natalie, a stay-at-home mom to their three kids, accompanied him. That Saturday, Jon, Natalie, and three of Jon’s co-workers met in the Marriott lobby, planning to head to a nearby bourbon distillery, where Strand’s company had made a dinner reservation. Someone ordered an Uber.
A few minutes later, Wallace pulled into the Marriott parking lot. The group piled in, but they had a last-minute request before leaving. Another co-worker wanted to ride along with them; could Wallace rearrange the back seat to accommodate a sixth person?
Jon Strand says Wallace protested, furiously. The request, Wallace argued, was for a group of five, and Uber’s app doesn’t allow for adding extra passengers to a ride already in progress. According to Strand, when the group offered to tip in cash for the extra passenger, Wallace accused them of trying to take advantage of him. The group hadn’t even left the hotel’s pickup spot when he ordered them all out of the car.
After this, Strand’s memory blurs. “I never saw the first punch coming,” he told me recently over the phone. Strand recalls blows to the face, blood on his arms, and pleading from his wife and co-workers as he and Wallace fought in the lot just outside the hotel. Then, according to the Strands, Wallace retreated to the car and fished out his gun, later determined to have been loaded with seven live rounds. One of the passengers broke away to call 911.
“We’ve got an Uber driver who’s got a gun,” says the 911 recording, obtained by The Atlantic. “We need help.”
Strand wrestled the gun from Wallace, tossed it away, and pinned him to the ground. “It wasn’t like Jackie Chan,” Strand told me. “It was rolling on the ground, hoping you don't get shot in the face and doing everything you can the whole time [to see] my little 5-year-old [again].”
As Strand restrained him, Wallace started to gasp. He told the group he couldn’t breathe. Strand’s co-worker placed a second 911 call. Wallace was in cardiac arrest, now lying limp in the Marriott's parking lot. Over speakerphone, an operator coached Strand and his co-workers on how to perform CPR as they waited for first responders to arrive.
Wallace died in the hospital three days later. He didn’t have life insurance through driving (Uber offers only third-party liability insurance, which isn’t the same as traditional employer-sponsored coverage), so a family friend created a GoFundMe campaign to support Wallace’s fiancée and kids. (Through their estate attorney, Wallace’s family declined an interview request.) In a GoFundMe update thanking donors, his fiancée wrote, “There are a lot of people, not just us, who loved Mike very much.”
As Wallace’s family was mourning, the Louisville Metro Police Department was investigating his death as a potential homicide. Though the autopsy determined that the cardiac arrest was triggered by the struggle, police ultimately found that Strand’s force that evening was justified. Police closed their investigation, and Strand returned to Indianapolis. But in a recent phone call, he described the abiding anxiety that started after the incident: flashbacks, nightmares, a parade of counseling sessions. He told me he was diagnosed with PTSD in the midst of the four-month homicide investigation, during which he had to retell his story again and again. Eventually, he left his job and devoted himself to his recovery.
In August, Strand found an outlet for his feelings: He filed suit against Uber, claiming that Wallace should have never been hired. According to court documents, in 2004, Wallace had pleaded guilty to menacing, a Class B misdemeanor that should have required him to forfeit any and all guns in his possession—and, the suit alleges, disqualified him from being hired by Uber in 2016. Strand’s lawsuit also argues that in promising a thorough background check of all drivers, Uber grossly misrepresented the safety of its product. Finally, and perhaps most consequentially, Strand’s lawsuit asserts that Michael Wallace was a full-time employee, not just a contractor, as Uber describes its drivers, making the company liable for his actions that night.
Strand's case is still pending, while his lawyers await a court date. Uber’s representatives declined multiple requests for comment on the record for this article, but, in an email to The Atlantic, they noted that the company has added new safety features to the app and updated its background-check system for drivers.
Wallace’s death and Strand’s lawsuit crystallize a central paradox of the gig economy. Since its inception, Uber (like its peers in the rideshare market) has held itself minimally responsible for the massive workforce of contractors that propelled it to a multibillion-dollar valuation—even as walkouts, lawsuits, and contested legal judgments have challenged this stance.
In Uber’s formal response to Strand’s suit, it not only denied that Wallace was an employee; it denied that any driver is. Uber “does not hire drivers and, therefore, there are no ‘Uber drivers’ and users of the Uber App are not ‘Uber Riders,’” its response to the suit reads. Rather, Uber calls itself an intermediary, a platform that connects people who need rides with independent contractors who drive cars. It refers to itself only as a “technology company” in the document and states, flatly, that it “does not perform any transportation service.”
But if drivers like Wallace are indeed contractors, Uber cannot override state laws in places like Kentucky, which give citizens the right to carry guns in their own cars. Strand’s suit argues that Uber’s self-perception is itself a danger to riders—one that both allows the company to shield itself from liability and prevents it from controlling for what happens when drivers disregard its rules. And, as it turns out, drivers do break the rules.
Uber has banned guns in cars, for both drivers and passengers, since 2015. But over email and Facebook Messenger, four current and four former drivers told me they carry firearms on the job. In explaining why, they each cited the same self-determinalist rhetoric Uber has slapped on subway ads to entice drivers and used in hearings to justify the business model: Drivers maintain good ratings, own their own cars, set their own hours, act as their own bosses, and follow local laws. But ultimately, they work for themselves, and Uber is, to use a Silicon Valley term of art, just a platform.
“Let them ‘ban’ guns,” a Las Vegas Uber driver told me. “They can’t enforce it.” Like many other drivers I spoke with, he told me he believes the company has effectively traded its right to enforcement for the cost savings associated with employing contractors: “Ultimately, I don’t think it’s Uber’s responsibility to make anyone feel safe,” he continued. “My safety is my responsibility. Uber is a software company. Their responsibility is the app.”
In 2017, Jose Mejia, a Miami driver, filed a federal class-action suit against Uber to reverse its firearm ban. Florida’s 2008 “bring your gun to work” law empowers employees to store legal firearms in personal lockers or their own cars. With Uber, of course, the car is the workplace. Mejia claimed that Uber policy violated Florida law and, citing an incident in which an Uber driver with a concealed-carry license shot and disarmed a Chicago gunman, argued that arming Uber drivers could save lives.
But Mejia couldn’t prove that Uber violated his rights: He hadn’t been fired or threatened with suspension. The company had announced a ban, yes, but never materially stopped him from carrying a firearm. The Florida court dismissed the suit without prejudice in 2018.
Here we have a uniquely American absurdity: Drivers can carry guns to work, to a bar, to a supermarket, but not in their own cars while using the app to transport passengers. Like Mejia, they exist in this space between name and effect, adherent to a ban with little practical enforcement.
“I follow all state laws,” said a driver in Utah, which, like Florida, permits drivers to carry loaded weapons in their own cars. “They can deactivate me, but in the event I need a firearm, that’s irrelevant to me staying on the platform.”
Only one of the drivers I spoke with, Marc D’Andre, told me he’d been deactivated or faced any enforcement from Uber for carrying a weapon. D’Andre, who is from Detroit, told me he was fired this summer after a passenger reported a handgun in his car. He says he protested, noting to Uber his concealed-carry license and clean driving record. His account, he told me, was still deactivated.
Like many other drivers, D’Andre noted that he puts himself at considerable risk—financial and physical—to drive for Uber, and needs a weapon to protect himself. Muggings and carjackings against Uber drivers are uncommon but not unheard of. None of the armed drivers I talked to had been attacked, but they each described a feeling of intense vulnerability, driving alone and often in areas they’re unfamiliar with, to locations unknown.
“[In] Detroit, just being in the wrong [area] at the wrong time of day or night in the right car could be a danger for you,” D’Andre said via Facebook Messenger. “Especially seeing that Uber requires you to have a newer car in good condition to drive for them, which heightens the possibility of being a target.”
Even when drivers are armed, they almost never use their weapons: “Safety incidents,” as Uber describes them, are rare. But Strand's case shows how guns can turn rare cases of conflict into catastrophes. “We’re dealing with real people, real family—not numbers, not percentages,” Strand said. “I’d want [Uber] to know the personal ramifications. If the customers aren’t safe, the results are devastating.”
Strand’s lawsuit follows the same argument used by labor advocates and researchers: that Uber’s system—which uses algorithms to manage drivers’ ride loads—is a form of control that should designate it as an employer. This clashes with how the company presents itself in legal filings: as merely a platform used by drivers who themselves provide transportation.
“There’s all these little ways in which [Uber] breaks down drivers’ true independence and stops them from making real, informed decisions about the jobs they take,” Alex Rosenblat, an ethnographer and researcher studying the gig economy, told me. Software monitors drivers’ every move, and the app nudges them to take certain rides. Until 2016, drivers could be suspended or terminated for a ride-acceptance rate below 80 percent. After a 2019 National Labor Relations Board memo noted that the company “limited drivers’ selection of trips [and] established fares,” Uber announced its Uber Pro service, which incentivizes high acceptance rates by offering extra rewards for drivers with higher ratings.
Earlier this month, California’s Assembly Bill 5 (AB5) went into effect; the legislation sets new standards about how companies designate contractors and employees. Uber is currently fighting the measure in court, calling it “irrational and unconstitutional,” but in the meantime the company has slowly begun testing features that offer more transparency for drivers, including giving them the ability to see a rider’s requested destination and how much they’d make from a specific trip before accepting it.
The bill’s author, California Assemblywoman Lorena Gonzalez, argues that Uber is hesitant to adopt reclassification in part because it would have to assume greater responsibility for the actions of its drivers than if they remained contractors. “Liability is one of the big unspoken-about issues here,” she told The Washington Post. In response, Uber spokesperson Noah Edwardsen told the Post that “liability for safety incidents has simply never been part of [Uber’s] arguments or strategy around AB5.”
By emphasizing driver autonomy, Uber has fine-tuned its service to avoid the exact liability that Strand is seeking. It’s true that many drivers pursue the job as secondary income, but many work 40 hours or more a week and use it as their primary family income. These are the people that Uber can assert enormous control over without issuing the direct mandates we associate with traditional employment.
The leverage Uber has over drivers—tacit, algorithmic, or otherwise—has limits. And when drivers decide that they need a weapon to defend themselves, the economic trade-offs between liability, classification, and autonomy seem tangential. In states where firearms are legally framed as the owner’s right, drivers need only follow the rhetorical path Uber has laid for them. Uber’s business model depends on driver “independence,” but stumbles when drivers choose to exercise it.