Uber launches data tasks as option for some U.S. drivers to earn money
Uber is seeking to ride the wave of a growing appetite for bespoke datasets and labeling services that require human vetting to train AI models.

Uber Technologies Inc. is giving some drivers in the U.S. the option to earn money by completing tasks related to the company’s nascent data labeling business, an area where the rideshare giant sees an opportunity to shine in the artificial intelligence boom.
A new job category called “digital tasks” will appear in the Uber driver app for some workers later this fall, offering existing registered drivers the opportunity to make a few bucks if they take on simple assignments that can be completed within minutes, according to the company.
Those stints currently include uploading documents like restaurant menus, or recording audio samples of themselves narrating a scenario in various languages, said chief product officer Sachin Kansal. More tasks will be added over time, he said, and the payout will vary based on the time commitment of each assignment.
Uber is seeking to ride the wave of a growing appetite for bespoke datasets and labeling services that require human vetting to train AI models. Scale AI, which offers similar services, received more than $14 billion in investment from Meta Platforms Inc. earlier this year and is valued at more than $29 billion.
The rideshare company’s data services business, Uber AI Solutions, has been pitching its services to other enterprises, inviting them to outsource some AI development to independent contractors. Late last year, it launched a web-based platform to recruit talent in more than 20 countries for new types of gigs like coding and translation.
Kansal said in an interview that the new digital tasks aren’t meant as an answer to potential driver displacement from the increasing availability of autonomous vehicles, but as a way to keep attracting drivers to the app with more ways to earn.
“It’s an interesting coincidence that this has been the year of us launching a lot of AV partnerships,” Uber AI Solutions and digital tasks for drivers, he said. “But if you look at the actual reality on the ground, we have hundreds of AVs. We have 8½ to 9 million drivers on the ground. So we are at a very, very, very early stage,” he added.
The company has been testing the new driver tasks in India ahead of the U.S. launch, Kansal said. Uber may consider opening up such gigs to nondrivers over time, he added, though the current priority is to increase the earnings of drivers and couriers already on the platform.
The digital gigs being made available on the driver app are designed to be simple enough for drivers to complete on their phones while they’re away from their cars, Uber said. Meanwhile, the assignments on its web-based platform have been more complex, such as annotation, translation, and editing multilingual and multimodal content.
Other updates to the driver app include expanding a feature allowing female drivers to set a preference for women riders. That option is now available in Baltimore, Minneapolis, Philadelphia, Seattle, D.C., and Portland, Ore. The company is also rolling out nationwide a feature that lets drivers set a minimum passenger rating for prospective riders.