Uber to Expand Gig Work for U.S. Drivers with AI Data Labeling

Uber (UBER) is looking to optimize its network and expand the earning potential of its drivers and couriers by introducing in-app micro-tasks, aiming to capitalize on downtime between rides and deliveries.

Announced at the Only on Uber 2025 conference in Washington, D.C., CEO Dara Khosrowshahi unveiled the new initiative, framed as a direct response to driver feedback gathered through extensive “Crew sessions” involving over 100 Uber team members. The company stated these sessions encompassed hundreds of hours, focusing on diverse aspects, from product design to policy changes.

The initial pilot program, spearheaded by Uber’s AI Solutions Group, will offer drivers the opportunity to complete small online jobs, such as uploading photos to train AI models. This concept, currently being tested in India, aligns with the growing trend of leveraging gig workers for AI development and data annotation. Another example provided by Uber spokesperson Meghan Casserly included recording spoken phrases in various languages and accents. Uber clarified that none of these tasks would relate to Uber’s partnerships or driverless vehicle development.

While Uber hasn’t disclosed specifics regarding project clients or the AI sectors involved, the move positions Uber as a potential player in the burgeoning micro-tasking market, competing with platforms like Amazon Mechanical Turk and Upwork. The implications for Uber’s data strategy are significant. By directly sourcing data annotation from its driver network, the company gains access to a vast and geographically diverse pool of contributors, potentially giving it a competitive edge in training its AI algorithms. The variable pay structure, contingent on task complexity and completion time, provides a flexible model for both Uber and its drivers.

This initiative also signals a potential shift in the gig economy. By offering more diverse income streams Uber could address concerns about earnings volatility and potentially increase driver retention. However, crucial aspects such as task availability, platform integration, and consistent payment models will determine the program’s success.

In addition to the new micro-tasking initiative, Uber announced the expanded availability of its women rider preference setting, designed to match women drivers and riders. Now available in Baltimore, Minneapolis, Philadelphia, Seattle, Portland, Oregon, and Washington, D.C., the service builds on prior implementations and underscores Uber’s commitment to safety and inclusivity. Khosrowshahi noted the popularity of the feature, citing over 100 million trips facilitated through the women driver option in markets where it’s actively supported.

Other driver-focused enhancements include the ability for drivers to set rider rating preferences and a “delayed ride guarantee.” The driver rating preference is a feature that allows drivers to avoid picking up riders with ratings below a certain threshold. The delayed ride guarantee addresses the issue of longer-than-estimated trips, ensuring drivers receive higher payouts for rides exceeding the projected duration by more than five minutes. While the company acknowledges that these additional costs may, in some instances, be passed on to the rider, the decision depends on the determining factor in the trip’s delay.