Senate Sleeps While ‘Bossware’ Continues to Surveil American Workers
Only three U.S. senators have signed on to the “Empowering App-Based Workers Act.”
By Steve Wishnia
A Senate bill to regulate employers’ use of software to monitor workers and algorithms to evaluate them and set their pay has won support from unions representing app-based workers, despite its apparently slim chances of passage.
“This bill is a critical step in the right direction,” Rideshare Drivers United President Nicole Moore said in a statement released by the National Employment Law Project (NELP).“Our members know that the algorithms behind these apps are basically 21st century robot bosses: they determine how much we earn, if we earn at all, and when we can earn. We need transparency and accountability.”
The Empowering App-Based Workers Act, introduced by Sens. Brian Schatz (D-Hawaii) and Chris Murphy (D-Conn.) on July 28, would require employers to give workers full notice about what monitoring and automated-decision systems are used to oversee them and what they are being used for, along with weekly pay statements with details about how pay is calculated and the prices consumers pay. It would also prohibit app-based cab companies from taking more than 25% of fares collected.
Violators could be fined up to $20,000 per violation, and $25,000 for violations involving whistleblowers. In civil actions, both the subcontractor and the company hiring them would be jointly liable. And third parties, including labor unions, could file lawsuits on behalf of workers.
“We’re supporting it,” New York Taxi Workers Alliance head Bhairavi Desai told Work-Bites. Particularly important, she said, is that it would bar companies such as Uber and Lyft from taking more than 25% of fares as its commission. Uber’s “take rate,” according to a Columbia Business School study released in June, can be as high as 50%.
Other supporters include the AFL-CIO, the Service Employees International Union, the United Food and Commercial Workers, the Hawai’i Workers Center, Los Deliveristas Unidos, and app-cab driver unions in six states.
The bill, however, has only three senators signed on as sponsors—Schatz, Murphy, and Tammy Baldwin (D-Wisc.). With Congress in recess in August, neither Schatz, Murphy, nor Senate Health, Employment, Labor, and Pensions committeechair Bill Cassidy (R-La.) commented to Work-Bites.
“It’s a stake in the ground,” NELP senior staff attorney Sally Dworak-Fisher told Work-Bites. It would prohibit “algorithmic wage discrimination” and “surveillance-based wage-setting,” she said, and create understanding about what these systems are. “It’s short of what we advocate, but it’s a start.”
Surveillance and control
NELP, in a report issued in July, defined “bossware” as “workplace digital surveillance technologies that businesses use to collect data on workers’ activities,” and “workplace automated decision systems that businesses use to assist in managing their existing workforce.”
Amazon, with its constant monitoring of workers’ pace and “time off task,” and Uber, with its tracking of drivers’ locations, are probably the best known, but it is widespread. The NELP report cited a 2024 survey that “showed that two-thirds of the workforce experience workplace digital surveillance and monitoring,” and one-third work under “automated management of tasks and schedules.”
One form is the “just-in-time” scheduling software used by retail and fast-food employers, which gives workers irregular, unpredictable shifts, keeping them on call for busy periods and sending them home early if business is slow. In nursing, gig-economy companies such as CareRev, Clipboard Health, ShiftKey, and ShiftMed use algorithms to hire temporary nurses and certified nursing assistants. On ShiftKey, they bid on available shifts by indicating the lowest wage they’ll take, according to “Uber for Nursing: How an AI-Powered Gig Model Is Threatening Health Care,” a December 2024 report from the Roosevelt Institute.
They do not know how much the hospitals or nursing homes they work at are paying the app company for their services.
Uber’s ‘upfront pricing’
With technology, “companies are able to divorce what it takes to pay a worker from what it takes to do the job,” says Dworak-Fisher.
That is particularly true at Uber. In 2022, the company switched from setting fares by the traditional per-mile, per-minute rate to “upfront pricing,” in which the customer is told the fare in advance. That fare is set by an algorithm that incorporates Uber’s “surge pricing” surcharges at places and times of high demand.
The Columbia Business School report called it the “largest-scale implementation of algorithmic price discrimination on both sides of its marketplace — enabling the company to raise rider fares and cut driver pay on billions of rideshare trips, systematically, selectively, and opaquely.”
The report’s analysis of tens of thousands of trips from 2019 to early 2025 found several ways that “Uber has exploited its opaque price and pay algorithms to increase profits at rider and driver expense.” They included reducing drivers’ base pay to lower their share of the surge-pricing bonus; raising the base fare on trips before offering passengers discounts; and charging more for trips between high-income neighborhoods.
“Upfront pricing” turned Uber from a money-losing operation into a profitable one, the report concluded. But “as Uber increased prices, driver pay remained static,” says Desai.
The company’s take rate increased from 32% to 42% of fares collected, the report said. According to its analysis of one driver’s 24,500 trips between 2019 and 2024, the share of trips where the price and the driver’s pay were within 10% of average fell from 86% to 22%. Trips where the price was more than 10% above average increased from 14% to 22%. And those where the driver’s pay was more than 10% below average went from zero to 34%.
On those low-paying trips, Uber’s take rate was 50%. Overall, on 28% of trips in 2024, the driver was paid less than he would have earned under the old rates.
“Upfront pricing gives riders clarity before they book and allows drivers to make informed decisions based on full visibility into pay, distance, and expected duration,” Uber said in a response included in the study. “Our dynamic pricing algorithms help balance real-time supply and demand to improve reliability across the platform.”
Drivers in New York have been somewhat insulated from these fluctuations, Desai says, because of recent Taxi and Limousine Commission regulations that set a minimum wage for drivers and limit what companies can take as a commission.
She believes the take rate should be limited to 15% in New York City, because the cost of mandatory auto insurance is highest here. Drivers are the ones who have to make the investment in owning and maintaining the vehicles, she emphasizes.
Personalized pay rates
A more sinister possibility is that employers are using personal data to try to predict who’s desperate enough to work for less—for example, Dworak-Fisher posits, a single parent who just took out a payday loan.
“We can’t be sure” how widespread it is, she says, as “they do everything to fight transparency.” But, she adds, “we have reason to think” that Uber and Lyft are using personalized data.
Uber has denied that. “Upfront prices are not personalized — our pricing algorithms do not use information about an individual rider or driver’s personal characteristics,” it told the Columbia report.
But companies already can analyze purchasing habits, urgency of need, and geographic constraints and adjust prices accordingly, Veena Dubal, the University of California-Irvine law professor who coined the term “algorithmic wage discrimination,” and Zephyr Teachout wrote in a February report titled “Prohibiting Surveillance Prices and Wages.” “The same technologies can be employed to set wages.”
“Uber and Lyft set compensation based on algorithmic determinations, and there is reason to think they use intimate data to determine precisely how little an individual worker may be willing to accept to perform a job,” it said.
Desai says there’s enough evidence of a spark, but “not enough to show a fire.” One way to test that theory, she adds, would be to examine records of trips between similar locations at similar times and see how much different drivers were paid.
Meanwhile, legislation to prohibit “surveillance pricing,” including wages, has been introduced in Colorado, Illinois, and Georgia, the NELP report noted.