Food delivery riders call on platforms to open up AI ‘black box’ they say has cut pay
Food delivery riders in Edinburgh are calling on platforms including Deliveroo, Uber Eats and Just Eat to reveal how their pay-setting algorithms work, arguing that growing use of automation has driven down earnings even as workloads have stayed the same. Riders working with the Workers' Observatory, a charity founded by gig economy workers and academics from St Andrews and Edinburgh universities, say the shift towards "dynamic pricing" has made pay unpredictable and left them unable to earn a stable living, adding to wider concerns about opaque AI systems governing gig work.
One rider, David, said he was earning half what he made four years ago for the same hours, while lead organiser Xabier Villares described a "dramatic change" in the last three years that has ended his ability to rely on the job to cover rent and bills. Research from Oxford University and Columbia Business School found Uber's 2023 dynamic pricing algorithm left drivers earning substantially less per hour, and this week drivers in the UK, Netherlands and elsewhere launched a class action in Amsterdam alleging the system breaches data protection law; Uber denies personalising fares based on driver behaviour. The Workers' Observatory, which has secured a decade of research funding, has run experiments such as one in Dunfermline where riders coordinated to reject low-paying job offers, temporarily raising rates for some.
- Edinburgh riders demand transparency on delivery platforms' pay algorithms.
- Workers say pay has halved in four years despite unchanged workload.
- Uber drivers separately launch Amsterdam class action over algorithmic pay.
New here? Start with this
Food delivery platforms such as Deliveroo, Uber Eats and Just Eat use computer systems to set the pay offered for each job. These systems can adjust rates as demand, location and other factors change, but riders say they cannot see how the calculations are made.
Many riders work as self-employed contractors and are paid per delivery, so changes in job rates can affect their income directly. The Workers’ Observatory, a research charity involving gig workers and university academics, studies how platform work affects people’s lives.
The issue matters because riders say unpredictable pay makes it harder to plan around rent and bills, even when they work similar hours. It also raises wider questions about how much information companies should provide when automated systems shape people’s earnings.
Both sides, in good faith
The strongest fair case each way — we don't pick a winner.
The case for
Riders argue that platforms should disclose how automated pay is calculated because opaque, changing rates make it difficult to predict income or challenge unfair outcomes. Their reported loss of earnings, alongside research suggesting dynamic pricing can reduce hourly pay, supports calls for transparency and accountability when algorithms shape people’s livelihoods.
The case against
Platforms could argue that dynamic pricing helps match delivery demand with available riders, keeps services operating across busy and quiet periods, and allows rates to respond to changing conditions. They may also say that revealing detailed pricing rules could make the system easier to game, while individual accounts and findings about drivers do not by themselves establish how rider pay is set or that the platforms’ systems are unlawful.