Gig workers in India unite to take back control of algorithms

On October 17, Santosh Kumar, an Uber driver from the southern Indian city of Hyderabad, was finishing a nearly 12-hour shift and struggling to find a final ride home. The app showed him a message that destinations in that area were unavailable.

Frustrated, he turned to a Telegram group called CCDA, or Commercial Cab Driver’s Awareness, where he shared his woes with more than 5,000 other drivers. Within minutes, his peers offered a jugaad — a cheap hack — to cheat the system: keep trying to book a ride home, and the algorithm will eventually oblige.

Two days later, in the same group, another distressed Uber driver posted screenshots of a “miscellaneous” fee of more than 5,000 rupees ($61) that Uber had taken from him. The screenshot stated that if he did not make the payment, he would lose access to his Uber account. He didn’t really understand how Uber calculated this amount and wondered how he could afford the hefty payment.

CCDA members explained that this was a mandatory tax payment and offered a jugaad to compensate for the single heavy load: keep accepting rides, and Uber will automatically deduct the amount from daily earnings rather than paying the large amount up front. “They explained to me that until this amount is settled, I will only get rides with online payments and not cash rides,” the driver said. Rest of the world, requesting anonymity fearing retaliation from Uber. He managed to clear over 2,000 rupees of tax debt in less than a month.

These are just two examples of how Indian gig workers – tired of the obscurity surrounding the algorithms and black box technologies that dictate their lives and work – are finding ways to leverage the platforms to their advantage. Drivers and delivery people, who work for apps like Uber, Ola, Zomato and Swiggy, are trying to reverse-engineer those apps, frequently sharing that information through groups like the CCDA and in-person workshops.

“Telegram groups are like our walkie-talkies,” said Shaik Salauddin, Uber driver and founder of Telangana Gig and Platform Workers’ Union (TGPWU). Rest of the world. Salauddin is part of 64 such telegram groups where Indian workers help each other navigate their lives working for tech platforms.

“The exchange of information is crucial because the drivers have very little information about their work: why they are paid; what they are paid; why they are associated with certain commands; why sometimes they get a lot of orders, sometimes they get none,” said Rida Qadri, who studies algorithmic failures and friction in the Global South, with a particular focus on Indonesia. Rest of the world. “The more they know, the more drivers feel they can master their job, generally earn more, stay safe, etc. – the exchange of information is then an important part of the driver’s job.”

The activity of these groups is reminiscent of similar efforts around the world. Last year, the Amsterdam District Court made a historical ruling in a case, brought by drivers in the UK and Portugal, asking Uber and Ola to provide more algorithmic transparency about the data they use. In Indonesia, Gojek workers have developed a series of gray market apps to fight against black box algorithms. One such app helps drivers spoof their GPS location to avoid waiting at a specific location to receive orders.

“Telegram groups are like our walkie-talkies.”

Although Indian workers have not gone so far as to create apps, they have developed their own mini-tech support system to help each other understand the algorithm dictating their jobs, developing tricks to make the system work .

“Workers know a lot of different ways – not about the algorithm or the variables that go into the algorithm – but they can tell you and speak fluently about surge pricing; they can tell you what’s going to happen at night or during rains or all these different rhythms of how the algorithm works,” said Srujana Katta, Ph.D. candidate at the Oxford Internet Institute, who is currently studying the work. carpooling and work organization in Hyderabad. “They are able to see the patterns even without knowing the logic behind the algorithm specifically.”

For example, Uber pushes drivers to go to places where the demand is higher and the supply of taxis lower. The app does this by displaying “surge pricing” on a heatmap. But several drivers said Rest of the world that over the years they realized it was just a ruse by the app to get them to a low supply location without giving them a higher rate. “The maps light up, turn red, and show the multiplied fare we can get if we go to that specific area,” said Kumar, who has worked at Uber for at least five years. “But as soon as we leave, the price hike is gone.” He said he stopped falling into this algorithm trap. Uber did not respond to an email from Rest of the world.

In the early 2010s, when Uber forayed into the Indian market, Salauddin recalls, the company lured drivers by offering incentives and hosting lavish biryani meals in upscale hotels. “They put up big billboards, invited drivers to big hotels, showed drivers a big dream that they could earn 80,000 [rupees] per month, 100,000 [rupees] per month,” recalls Salauddin, who is also the National General Secretary of the Indian Federation of Application-Based Transport Workers, or IFAT. But this dream was short-lived. “A few months later, we could already see things going south. As more and more people attached themselves to these platforms, revenues started to drop,” he said.

Selvaprakash Lakshmanan

To alleviate this growing pressure, IFAT and TGPWU constantly keep in touch with gig workers across India and hold different types of workshops including those to educate them on algorithms and data protection, among others.

“Many [drivers] are not educated enough,” said Salauddin Rest of the world during a June visit to his office in Hyderabad. “So we tell them how these apps and algorithms work, how to protect their data, how to change their phone settings and not give full access, and some other little tricks that can help them.”

These workshops can be about something as basic as an introduction to Telegram, so gig workers can stay in touch with each other and keep up to date with the latest tech hacks. Salauddin said the TGPWU held more than 40 such workshops over three months in early 2021.

Last year, TGPWU also held workshops for taxi drivers and delivery people to help them understand how algorithms calculate their income. “Until a year ago, they didn’t understand how and why the platform was deducting the money,” Salauddin said. “Now they understand and publish these calculations on Telegram groups.”

In April, the Bengaluru-based NGO IT for Change held a data protection awareness workshop. “They run workshops and train us, which we then disseminate to others,” said Salauddin, who was invited to attend the workshop. “We explain to our members the value of our data; we show them how to have their settings to protect their data because, when we install these apps, we give access to everything, right? So we teach them how to access the settings and check where to give access or not. »

While gig workers haven’t found workarounds to all of their problems with these apps, the first step towards this is awareness. Salauddin said unions are also trying to figure out how the algorithms of super-fast grocery delivery platforms work. “How the algorithm decides who gets what order, how many orders, what’s the target – we’ll have to figure it out because some get 10 [orders], some get 15; for some it’s 20,” Salauddin said. “The delivery people are also confused.”

In the near future, IFAT has big plans to tackle algorithms that “manipulate” gig work, said Prashant Sawardekar, IFAT’s president. Rest of the world. His biggest gripe with ride-sharing platforms is that their algorithms favor taxis going through their network of providers, even as independently driven vehicles jostle for rides. “These cabs [through vendor networks] are dirty; the drivers don’t even have high marks but they are the ones who keep getting driven,” Sawardekar, who is based in Mumbai, said. Rest of the world. “The independents don’t get enough rides, even though they have high ratings. They [cab-hailing platforms] say reservations are given by their system, so how is that possible? »

Sawardekar said the union plans to take the matter to court.

Meanwhile, Salauddin – with help from IT for Change and the International Transport Workers’ Federation – plans to publish a detailed report soon on how algorithms are manipulating workers and their earnings in India.

“This report is about how platforms steal workers’ income, how they steal our data; we want to create awareness and for this to be in the public domain – it is heavily focused on revenue calculations,” Salauddin said. “I want to spread this report everywhere and start a campaign. We also plan to ask drivers to have this report in their taxi, to inform customers of the discrepancy between what they pay and what we actually get.

Reporting of this story was supported by the Pulitzer Center’s AI Accountability Network.

Sharon D. Cole