New Research Exposes Deepening Exploitation of Uber Drivers by Algorithmic Pay

A new independent study, published today by researchers from the University of Oxford in partnership with Worker Info Exchange, reveals that Uber’s introduction of dynamic pay has reduced driver pay, increased Uber’s commission, and made pay more unpredictable and unequal for its workforce. Uber’s so-called ‘dynamic’ pay system uses artificial intelligence and machine learning to set pay and assign work in real time — with no transparency for drivers and no ability to challenge automated decisions for pay and work allocation.

The report — Not Even Nice Work If You Can Get It; A Longitudinal Study of Uber’s Algorithmic Pay and Pricing — is based on an algorithmic audit of 1.5 million trips from 258 drivers who accessed their own trip data with Worker Info Exchange using GDPR subject access rights. This participatory audit, co-designed with Uber drivers themselves, is the first of its kind to leverage data subject access requests (DSARs) at scale to evaluate the impact of algorithmic pricing on gig workers’ livelihoods.

Key Findings

• Pay has declined since the introduction of dynamic pay. Gross hourly pay dropped from £22.20 to £19.06 before considering operating costs including vehicle rent, maintenance, fuel, commercial insurance, licensing, cleaning, parking, congestion charging and other costs. This leaves many drivers consistently earning below the national minimum wage despite the notional protection of the 2021 Supreme Court ruling against Uber.
• Pay inequality has widened, indicating active algorithmic wage discrimination. 82% of longer serving Uber drivers are now earning less per hour than they did before dynamic pay and pricing was introduced. A small minority are earning more, but these gains are not shared evenly with benefits going to newer drivers and part time workers.
• Pay has become less predictable. Machine learning models show that drivers’ ability to anticipate earnings based on time, location, and trip type has drastically diminished. Whilst Uber often claims that drivers are free to work where and when they want, this report shows that the tacit knowledge of working drivers on the job has been degraded by Uber’s algorithms who seek to shape the market to their advantage only.
• Uber’s commission (“take rate”) has increased and become less transparent. Prior to dynamic pricing, Uber’s commission was fixed at 25%. After the introduction of AI driven dynamic pay, Uber is taking much higher cuts — up to 50% or more on certain trips.
• Drivers’ working time is increasingly unpaid. Drivers now spend more time on standby waiting for dispatches than on paid trips. Standby hours have risen by over 1 hour per week on average since 2022. WIE’s previous Dying for Data report estimated the cost to UK workers for wage theft associated with unpaid waiting time was £1.9 billion in 2023.

A Shift Toward “Algorithmic Gamblification”

The study characterizes Uber’s pricing shift as a form of “algorithmic gamblification” — where drivers are pushed into a real-time gamble for fair pay, with no transparency, no bargaining power, and no control over how prices are calculated.

As one driver put it: “They [Uber] are robbing us and the customer.”

How much have UK Uber drivers lost out in pay?

Building on the University of Oxford’s analysis, Worker Info Exchange has conducted its own assessment to estimate total losses for Uber drivers in the UK. According to Uber’s financial disclosures, the company’s global take rate — the percentage of fare revenue retained by Uber — increased from 20% in Q4 2021 to 30% in Q4 2024, following the introduction of dynamic pay. This 10-percentage point increase represents a $8.7 billion global loss of income for Uber drivers in the 12 months to March 31, 2025. Given that Uber UK accounts for approximately 18% of global revenue, Worker Info Exchange estimates that UK Uber drivers lost out on $1.6 billion in pay over the same period — a direct result of Uber increasing its share of the fare.

Worker Info Exchange Urges Urgent Policy Action:

• Mandate transparency on pay and pricing algorithms for both drivers and passengers.
• Ban dynamic pay and end to unfair automated decision making with effective enforcement of the UK GDPR.
• A single status of employment to end bogus elf employment and active enforcement of the 2021 Uber Supreme Court ruling to which Uber has never fully complied with.
• An end to wage theft with pay for all working time — from log on to log off — is paid, as ruled by the Employment Tribunal against Uber.
• Cap private hire licensing of drivers to improve driver utilisation, ensure fair pay and reduce congestion.
• Give unions collective data access rights and provide anonymised journey data to local authorities for open publication — as already happens in New York, Chicago, and California.

James Farrar, Director of Worker Info Exchange, said:

“Uber UK managers must now come clean and explain to their workers how their pay is set — and how much of each fare the company is taking. If Uber is allowed to continue getting away with the algorithmic trickery of its so-called ‘dynamic pay’ model, we should not be surprised when hyper-variability and AI-induced precarity in pay become the norm across the entire labour market.”

Reuben Binns, Associate Professor of Human Centred Computing, University of Oxford said:

“Uber’s dynamic pricing and pay algorithms are squeezing drivers while fleecing customers. The more customers pay, the higher Uber’s cut, and the less drivers keep, meaning that drivers are worse off even as they generate more profits for Uber.”

In June of last year, the IER published a report on Algorithmic Management, ‘Algorithmic Management and a New Generation of Rights at Work’, authored by Dr Joe Atkinson and Prof Philippa Collins. In it, the authors argue that existing worker protections, already hugely undermined by decades of anti-union legislation, are woefully inadequate to meet the challenges posed by these developments. They contend that a new regulatory framework is urgently needed, giving workers and their unions a genuine voice in the use of algorithmic management tools, alongside recognition and protection of their rights and access to justice that ensures their employers can be held to account.

You can access the report here.