Drivers Protest Uber’s ‘Black Box’ Fare System

Uber’s use of a secretive fare pricing algorithm driven by artificial intelligence lowers drivers’ wages, causes them confusion and uncertainty, and could undermine public safety — all while boosting company profits to record levels. 

That’s the conclusion of a new report surveying more than 2,500 Uber drivers and the message sounded by dozens who protested July 1 outside the company’s San Francisco headquarters. 

Drivers say the company’s opaque fare pricing system has been hurting their income and wellbeing. The protesters’ signs echoed conclusions from the report by PowerSwitch Action and Gig Workers Rising, grassroots organizations representing workers and labor interests.

Uber’s AI-driven algorithmic fare pricing, first rolled out in July 2022 as “upfront pricing,” draws upon huge volumes of data gathered from drivers and customers who use its app. The system can offer different drivers fluctuating, unpredictable fares for identical customer rides.

Drivers have done independent experiments — putting their phones on one table with the Uber app activated — that show the company offers them the same customer rides at different fares, meaning one driver would earn less than another for driving the exact same route. Drivers have done the same experiment with the Lyft app with similar results.

Experts call these discrepancies algorithmic wage discrimination, because the apps take into account the history of individual drivers’ activity to determine how little a particular driver is willing to get paid to pick up a ride.

In a February 2024 earnings call with investors, Uber CEO Dara Khosrowshahi touted the benefits of the company’s fare pricing approach: “What we can do better is actually targeting of different trips to different drivers based on their preferences or based on behavioral patterns that they’re showing us.” 

Uber disputes the criticisms and conclusions drawn from the driver survey. Spokesperson Zahid Arab wrote in an email: “This one-sided narrative is out of touch with how the Uber platform really works. Drivers aren’t punished for turning down trips — they have full control, and Upfront Pricing lets them decide what’s worth their time. The report’s methodology is vague, the sample is unrepresentative, and the conclusions seem driven more by agenda than by evidence.” 

Prior to the arrival of upfront pricing, driver compensation had some variability based on driver tenure and ride dispatch times. But it guaranteed a trip fare that was based on per-mile and per-minute rates. Now, many drivers say the opaque system results in lower pay for drivers and higher profits for the company.

“Uber now only has to pay the minimum any driver will accept for each trip,” Columbia Business School Professor Len Sherman wrote in a Medium story cited by the report.

Mariah Montgomery, national campaign director at PowerSwitch Action, said drivers are frustrated by the lack of transparency in Uber’s algorithmic fare pricing.

“We have a black box,” she said. “They don’t know what’s going on inside there. They just know that what they’re getting is not equal pay for equal work.”

Montgomery said the uncertainty of the system is confusing. 

“Drivers don’t know how the app may be manipulating their pay and the need for transparency is part of it,” she said. “But more than that, we just want all drivers to have dignity in the work that they do, and safety and predictability and security.”

The report noted that since introducing algorithmic pricing, Uber has seen a huge jump in its share price, which closed Wednesday at $96.64, up from $23.18 in July 2022 when upfront pricing was first announced.

In 2023, the first full year with upfront pricing, Sherman wrote that average monthly income for ride-hailing drivers declined 17% from the previous year “a uniquely steep pay cut compared to its gig company competitors Lyft, DoorDash, Instacart, and Grubhub.”

He also found, based on data from Gridwise Analytics, YipitData and other sources, that Uber’s “take rate” — the difference between the fare a rider pays and what the driver is paid — had spiked to 40% by the third quarter of 2023, crediting AI-driven algorithmic pricing for the company’s suddenly increased profitability.

Uber disputes Sherman’s findings on take rates; Arab called them “wildly and historically misleading.”

But some of Sherman’s findings were supported by an earlier study, by researchers at the University of Oxford, that was covered in a June 2025 story in The Guardian. The researchers reported that Uber’s fare cut in the UK ranged from 29% to more than 50% per ride after it switched from its previous practice of taking a fixed percentage of each fare. The company’s previous high take rate there was fixed at 25%.

Using AI to squeeze drivers

The PowerSwitch Action and Gig Workers Rising survey shows 73% of drivers who declined low-fare rides reported that the Uber app offered them fewer rides or continued to offer them only low-fare rides. One-third of drivers reported that when they were approaching the number of rides needed to score a bonus, the Uber app slowed down ride offers, making it harder to receive their bonus.

According to the report, drivers said they were being gamed by the company, with 78% responding “that driving on the Uber app feels like gambling — the occasional good fare keeps them going.”

Uber disputes claims that the app slows down offers to drivers who decline low-fare rides or that factors determining fare prices are not transparent.

The gambling-like feelings that drivers report may not be coincidental, according to Veena Dubal, a professor at the School of Law at the University of California, Irvine. In her study, “The House Always Wins: The Algorithmic Gamblification of Work,” Dubal described drivers’ feelings of being constantly tricked by the app as “ride-hail roulette” and that this psychological manipulation seems to be by design.

Uber rejected suggestions in the driver survey report that fares were personalized or manipulated. However, reporting by the Los Angeles Times — which revealed potential wage discrimination by several companies that hire gig workers — scholarly research by Dubal and the side-by-side experiments done by drivers seem to show otherwise.

An April 2017 story from The New York Times based on interviews with company insiders described how Uber’s social and data scientists experimented with low-reward incentive techniques gleaned from video games to keep drivers working harder and longer.

Uber leverages drivers’ desires to achieve earnings goals by offering them incentives to continue driving and sending them alerts about hitting targets. The most recent version of the company’s app offers up a next-customer fare while a driver is completing a current fare, leaving little time to make a decision, lest they risk losing the next one.

Drivers say these pressure-driven incentives compromise public safety by distracting them. According to the National Highway Traffic and Safety Administration, distracted driving is a major cause of traffic collisions, claiming 3,275 lives in 2023.

Uber drivers report concerns

For their survey of drivers, PowerSwitch Action and Gig Working Rising heard from 2,552 Uber drivers nationwide who responded to ads on Facebook. To be eligible, drivers had to be U.S. residents, have driven at least 10 hours per week for the last three months and correctly answer “driver knowledge” questions.

Uber disputes the validity of the survey, pointing out that it did not adequately account for part-timers who drive to supplement their incomes, and who Uber says represent the majority of its drivers.

The survey found 72% of drivers reported that in the last three months, it was more difficult for them to earn the same amount of money they did a year ago on the Uber app. More than half of drivers reported that in the last month, they earned less in a day than they expected to while using the app several times a week.

Marc, a veteran Uber driver who asked that his last name not be used, said he had completed more than 24,000 rides and watched his Uber income drop 20% annually for the last few years. “It reached the point where I had fallen below the line of being able to earn enough in a day’s work for 10 to 12 hours,” he said.

The report also noted that the new algorithm encourages drivers to stay on the road longer, which could mean more of them driving while tired and presenting a public safety risk.

The push to drive while tired is more pronounced for the 45% who say they drive for “essential income.” The AAA Foundation for Traffic Safety found that about 18% of fatal car crashes from 2017 to 2021 involved drowsy drivers. Lack of control over earnings also affects drivers’ mental health, with 68% reporting one or more measures of psychological distress in the last month.

The driver survey reported that the negative effects of Uber’s AI-driven algorithmic fare pricing were heightened for drivers of color; 80% of that demographic group reported being either overtime drivers or essential income drivers, compared with 68% of white drivers.

Algorithmic wage setting is starting to appear in other industries, including on-demand nursing and even for traditional W-2 paycheck employees.

While California is taking steps to regulate some forms of algorithmic decision making, so far there is little legislation that would address its effects on wages. PowerSwitch Action and Gig Workers Rising in their report made 10 recommendations for policymakers to improve the situation for Uber drivers and gig workers, including prohibiting continuous mass surveillance of workers (i.e., gathering data about them that isn’t essential for their job function), banning algorithmic wage discrimination and protecting against unfair terminations based solely on AI algorithmic decision making.

Drivers may get some relief through union protections. California lawmakers are once again considering legislation that would give drivers for ride-hailing companies a right to unionize with Assembly Bill 1340. That follows on California voters in 2020 approving Proposition 22, which exempted Uber and Lyft from classifying drivers as employees.