Uber rolled out an AI-powered pay model this week in Ontario, a change drivers worry will cost them income — and consumer advocates say may lead to higher prices for passengers.
And the change, which launched in B.C. in September, is raising questions in the industry about how ride-hailing and delivery companies use the data they collect as a form of power over their workers.
“Uber has found a way to take more money from us,” said George Wedge, an Uber driver in Toronto.
Compensation used to be based primarily on the time and distance of each ride and, in part, surge pricing, which was driven entirely by demand. The new system, however, uses an AI algorithm to calculate a driver’s pay using a host of other factors, like where a rider is picked up or dropped off or the day of the week.
The “upfront offers” system, which Uber has been using in the U.S. for two years, still includes surge pricing.
A new algorithm that determines how much drivers get paid for their rides is upsetting some drivers and consumer advocates who say it will lower driver wages while increasing rider prices. Uber Canada says the new system offers drivers greater ‘transparency and choice’ about whether or not to accept a trip.
How ‘upfront offers’ work
With the new system, Uber says drivers will see how much they’ll be paid before they accept a ride.
Keerthana Rang, a spokesperson for Uber Canada, said the change was made based on driver feedback and offers them “more transparency and choice when deciding whether to take a trip.”
It also gives drivers “a detailed breakdown of where the money customers were charged went,” including third-party charges, government fees, taxes and the company’s service fee, which will typically be lower for longer trips and higher for shorter trips, according to the statement Rang sent to CBC News.
But Wedge, who is also president of Rideshare Drivers Association of Ontario, says the new system may be transparent about how much someone will be paid — but it’s not clear how Uber determines the value of the factors that go into that calculation.
“It’s based on something that’s absolutely unknown to us,” he said, saying he worries about how the data collected from each ride will be used.
And since the system launched in B.C., Wedge said drivers there have told him they’re making less money than before the change came into effect.
In a December 2023 online post, Uber said driver earnings “have grown nearly 30 per cent in the last six years, faster than inflation.”
How will it affect riders?
Passengers advocates say they’re worried the new pay model could be bad for riders.
Brendan Agnew-Iler is the co-founder of the Toronto based RideFair Coalition, a not-for-profit focused on equitable and sustainable urban transportation.
Uber’s new algorithm is “taking an unfair system and making it vastly worse,” he said.
He said he believes Uber intends to offer the minimum amount that will lure drivers to work for them — and then charge the maximum amount the algorithm tells them passengers will tolerate, comparing it to how prices are set by airlines or for concert tickets.
“The public knows … that with dynamic pricing they will charge people whatever they think they can afford,” Agnew-Iler said, “especially in a monopoly situation, which is exactly what we have here.”
Toronto Metropolitan University economist Viet Vu isn’t sure the new algorithm will lead to significantly higher prices for riders — and he says there could be an advantage: It could take less time to hail an Uber.
Vu, the manager of economic research at the public policy think-tank Dais, said the new system could attract more drivers in peak demand times by offering better pay, “so if anything, the wait time will likely go down.”
Algorithmic pay and possible wage discrimination
In the U.S., the AI-based pay models used by both Uber and Lyft have been criticized as a form of discrimination by Veena Dubal, an American professor who studies gig work.
With algorithmic wage discrimination, two different workers can be “paid different hourly wages — calculated with ever-changing formulas using granular data on location, individual behaviour, demand, supply, or other factors — for broadly similar work,” Dubal writes in a paper published in the Columbia Law Review last year.
Her theory was tested in a video by the Rideshare Guy, a former Uber and Lyft driver, and another video by More Perfect Union, a non-profit news organization, with each showing drivers getting different pay rates for the same ride requests.
Both Lyft and Uber denied their pay systems are discriminatory in statements to NPR in 2023, and Lyft did not respond to CBC News about how they pay Canadian drivers in time for publication.
What about minimum wage laws?
The new pay model could be affected by recent legislation adopted in B.C. and labour legislation that will come into force in Ontario next year, both of which are meant to give gig workers some degree of wage security.
B.C.’s minimum wage law for gig workers, like food delivery couriers and ride-hailing drivers, guarantees they’ll be paid at least $20.88 per hour — but it doesn’t apply to the waiting periods between trips.
Ontario’s Digital Platform Workers’ Rights Act, which comes into effect in July 2025, will set a minimum wage for each assignment, mandate that there be transparency around how work is assigned — and the right to know how compensation is calculated, something that could be an issue with Uber’s algorithm.
Uber says it will comply with the law, and Vu says how much workers learn about how their pay is determined will be key to them figuring out when to work.
“That formula is kind of a secret,” said Vu, “so that leaves very little room for drivers to actually understand and sort of strategize their work.”