Food delivery riders across the UK are sounding the alarm over what they describe as a systematic erosion of their earnings, pointing the finger at the growing role of artificial intelligence in determining how much they get paid. Workers in Edinburgh have joined forces with academics to demand greater transparency from the “black box” algorithms that control their livelihoods, as platforms like Deliveroo, Uber Eats, and Just Eat increasingly rely on automated systems to match jobs with riders and set prices.
The Algorithm Problem: How AI Is Reshaping Pay
David has spent seven years navigating Edinburgh’s streets on his bike, delivering meals to hungry customers. His earnings tell a stark story. “I am making half the money I was making four years ago, for the same amount of hours,” he says. “It makes no sense.” His frustration is shared across the gig economy, where workers say the introduction of dynamic pricing algorithms has created a system where earnings become increasingly unpredictable and constantly squeezed.
The Workers’ Observatory, a charity established by delivery riders alongside academics from St Andrews and Edinburgh universities, has been documenting these changes. The organisation recently secured research funding that will sustain its work for the next decade, allowing it to dig deeper into the mechanics of platform operations.
Xabier Villares, the observatory’s lead organiser, has been pedalling for eight years. “There has been a dramatic change in the last three years,” he explains. “I used to work four or five days a week, especially evenings and some lunchtimes,, and was able to pay my rent and bills and make a decent living with that. But that’s not an option any more.”
Documenting the Decline: Rider Data Exposes Systematic Undercutting
Some riders have taken it upon themselves to track exactly what’s happening to their pay packets. Dylan has ridden for Deliveroo in Scotland for more than five years. Since mid-2023, he has maintained detailed records of his monthly income, the number of orders delivered, and his average fee per delivery.
His data reveals a consistent pattern. The number of orders Dylan completes each hour has remained steady at between 3.6 and 3.8 for the past three years. Yet his average fee per order tells a different tale. It has dropped steadily from £3.67 in 2023, to £3.63 in 2024, £3.51 in 2025, and £3.42 in the first half of 2026. “I started noticing a couple of years ago there was actually a big drop-off in my earnings,” Dylan explains. “If you are doing the same amount of work, that equates to getting paid quite a lot less for the same work.”
He attributes this decline to the algorithm constantly pushing rates downwards, along with the introduction of “stacked orders” by Deliveroo, where riders can accept multiple deliveries from nearby restaurants but receive fees that don’t match what they would earn for separate trips. “Some of the offers they give you view you as a clown,” he says.
Graeme Frances, another Edinburgh rider, believes the platforms “rely on this work being opaque”. He adds: “You don’t know what is in the black box of how they work out what to pay to whom.”
Facial Recognition Fails and Platform Accountability
The increased automation extends beyond pay calculations. Frances found himself locked out of the Deliveroo app after an accident left him with a black eye in August 2025. The platform’s facial recognition technology failed to recognise him, preventing him from logging on and earning money. He was unable to access human customer support to resolve the issue.
Deliveroo confirmed that Frances failed multiple facial recognition checks due to a blurry picture, though the company states he would have been able to work after a manual review, which he eventually passed. The incident highlights concerns about automated decision-making that can instantly cut off workers’ access to income.
Cailean Gallagher, director of the Workers’ Observatory and a lecturer at St Andrews University Business School, describes how “there is so much infrastructure of knowledge and data that’s concealed” by the platforms, leaving riders “working in the dark”. His team is attempting to change that dynamic through direct research.
Global Legal Challenges and Industry Response
The Edinburgh riders’ concerns echo across the industry. Research from the University of Oxford and Columbia Business School found that Uber drivers earned “substantially less” per hour following the introduction of a dynamic pricing algorithm in 2023. This week, drivers from the UK, Netherlands, and other countries launched a landmark class action against Uber in Amsterdam, alleging the company’s AI-powered system breaches data protection laws and systematically reduces earnings. The legal case could involve billions of dollars.
Deliveroo, which was acquired by its American rival DoorDash in 2025, has named its algorithm “Frank” and describes it as “super smart” technology that predicts order timings and determines which rider receives which job. The company insists Frank uses machine learning to deliver food “in the most efficient and reliable way”.
All three major platforms maintain that their riders earn above the national living wage while actively completing deliveries, though this calculation excludes time spent waiting for job offers. A Deliveroo spokesperson highlighted that the guaranteed minimum hourly fee “rose by 3.8% this year – an above inflation increase” and noted ongoing partnerships with the GMB Union to improve standards.
Why it Matters
The struggle unfolding in Edinburgh represents a pivotal moment for the future of work. Millions of people across the UK now rely on gig economy platforms for their income, yet the algorithms controlling their earnings remain largely opaque and unaccountable. As artificial intelligence plays an increasingly central role in determining pay and working conditions, the outcome of this battle for transparency could set precedents that extend far beyond food delivery. If companies can systematically reduce worker earnings through algorithmic adjustments without meaningful oversight, the implications for labour rights, income security, and the fundamental nature of employment in the digital age become profound. The Workers’ Observatory’s decade-long research commitment suggests this is just the beginning of a long campaign to bring fairness and accountability to an industry that has operated too long in the shadows.
