Michael Vi/Shutterstock
Artificial intelligence has already thoroughly infected the auto industry, whether we're talking about automakers that want to charge you more with AI, companies that directly advertise to an AI audience, or the Dutch relying on AI to ticket cars. Rideshare companies have gotten in on the action as well, and it's not just Uber trying to turn its driver's cars into AI-training data gatherers.
Along with its rival Lyft, Uber also has turned to AI to help set pricing — and we'll turn to a recent Consumer Reports study to help find out how it works. Just keep in mind that CR editors and the rideshare reps have vastly different ideas about both what the findings mean and CR's methodology. For example, CR had its volunteers check on pricing for the same routes at basically the same time – "generally within six minutes of one another and, in many cases, within the same minute." Nonetheless, the overall median difference in the study between the highest prices quoted and the lowest for the same journey at the same time was 42.4%. Uber's response was that it set fees based on market conditions during the exact moment the company is contacted. So you can't really compare rides ordered even seconds apart due to potential changes in supply, demand, etc.
True, CR admits that it didn't take into account a lot of factors that go into the pricing decisions, including driver supply, traffic changes, and more, but that's because the rideshare companies keep that kind of info themselves. Plus, CR didn't believe that they made a difference in the findings anyway.
How Uber and Lyft say they set prices
1000 Words/Shutterstock
This isn't the place for a deep dive into how supply and demand affect how much products and services cost. But rideshare companies do claim that changes in those conditions are what explain any differences in rider fees. CR isn't so sure.
According to the study, Uber and Lyft seem to be charging different prices based at least partly on factors that aren't related to economics. What can explain the differences? CR points to the incredible amount of personal data, including your past rideshare history, that the likes of Uber and Lyft can collect from your phone. As mentioned, rideshare companies don't share what kind of info they're extracting from people. So the public has to take their word for it that fares aren't based on gender, income, race, and similar factors – even when it seems like Uber/Lyft have the technology to discover that information.
For instance, Uber has a patent for tech that can access a phone's sensor data. That, in turn, can then be used to analyze a person's walking speed and style. The thing is, the same information can also help determine someone's gender. As a result, Uber could be charging higher prices for women. Your rideshare history can tell a lot about you as well. Uber itself gives an example of how it can identify a single working parent from the destinations that person is riding to. Would the company charge more to someone with a history of riding to the local HQ of the Democratic Socialists? The bottom line is that CR found no concrete evidence that they do, but positive evidence that they can.
Increased profits and fake discounts?
Diy13/Getty Images
Although the CR study was designed to look at how Uber and Lyft use AI to get more money from customers, it also raised a couple of related concerns. For instance, the same month-long investigation found the companies were taking bigger and bigger bites out of the ride fares. Per CR, between 43% and 49.5% of each fare goes to the companies – with the drivers' shares falling noticeably in the past few years. The typical take for Uber, for example, was 32% in 2022 and had jumped to 42% at the close of 2024. CR essentially connects the increase, and Uber's resulting profits, to the increased use of algorithmically generated pricing. On the other hand, correlation is not causation. You can decide for yourself.
As for discounts, these are common at both Uber and Lyft, and it looks like they're becoming an increasingly important part of both companies' playbooks. However, despite more denials from Uber and Lyft, CR reported that the companies regularly offered "supposed discounts on what appeared to be inflated original prices." Consider a situation from the CR investigation: One person was quoted a price of $82.08 for a trip, but with a discount that dropped that cost to $65.95. Yet another 40 riders were all quoted prices of just over $65 for the same route. Per CR, that means "the 'discount' from $82.08 is fake. $65.95 is the route's approximate starting price."
Oh, and CR further reports that both Uber and Lyft have admitted to using a customers' personal data for these discounts. No one explains how that's different from using personal data to set fares, however.