Uber trims leadership in its AI data-labeling push, while saying the unit has momentum

Uber has made two leadership exits tied to its emerging AI data-labeling unit, according to a report carried by Yahoo Finance. The change involves the removal of senior director Naga Kasu and product director Pankaj Kamat from the group responsible for AI-related data work, which is a critical early step in training machine-learning models.

The report, which was described as first appearing on GuruFocus, frames the transition as an internal management move rather than a shutdown. Uber reportedly said the AI effort is still seeing strong momentum, even as it restructured who oversees the work.

The term “data labeling” refers to the process of attaching categories and annotations to raw information so that algorithms can learn patterns. For AI systems that interpret images, text, or location-based indicates, labeling is often labor-intensive and can become a bottleneck if teams do not scale or coordinate it effectively.

While the leadership exits are the central development, the reporting also points to the broader organizational reality that AI spending and execution are moving targets for many large tech-enabled companies. In a separate, recent Fortune report on Uber’s AI cost scrutiny, the company’s COO questioned how to draw a line between rising AI expenses and customer-visible features, reflecting internal pressure to justify spend.

For Uber, the challenge is less about whether AI can improve experiences, and more about execution, including how quickly data pipelines can be built and scaled, and how teams measure whether training and labeling efforts translate into products riders actually use. A leadership reset can be a way to change priorities, restructure workflows, or refocus metrics.

Uber’s AI investments are also happening in a sector where cost control has become a competitive issue. AI initiatives can require sustained spending on computing, tooling, and operational labor such as labeling and review. That combination can raise the stakes for management teams tasked with both building models and demonstrating results.

The company did not provide additional detail in the coverage about why the two leaders were removed or what changes will follow, such as whether responsibilities will be consolidated, how the unit’s roadmap may shift, or how staffing will look next. The reports also did not disclose any immediate metrics tied to the unit’s performance, such as labeling throughput, model accuracy improvements, or adoption across Uber products.