Uber Optimizes AI: Business Models and Cost Management in 2023

Uber Leverages AI for Increased Efficiency

Uber is diversifying its use of artificial intelligence in its daily operations, according to CEO Dara Khosrowshahi. The company has managed to keep its AI spending "overall stable" during the second quarter, as noted by its CFO. To achieve this, Uber is utilizing more cost-effective AI models and encouraging its employees to monitor their expenses. Rather than opting for radical changes, the company prefers to implement numerous small improvements through AI, Khosrowshahi explained.

Expense control regarding tokens has been a major focus for Uber during the second quarter. The company has optimized its default settings for various use cases, adopted less expensive models for specific tasks, and encouraged its employees to better understand and manage their spending, Balaji Krishnamurthy explained in a statement following the quarterly results.

"As a result, the cost per token has decreased over the past few months, even as adoption continued to increase, allowing us to maintain overall stable AI spending," Krishnamurthy added.

Budget Management and Technological Innovations

Earlier this year, Chief Technology Officer Praveen Neppalli Naga made headlines by revealing that Uber had exhausted its Claude Code budget for 2026 in just a few months. Currently, Uber is using AI to make numerous small adjustments to its app and services, rather than making major decisions. For instance, the Uber app offers destination suggestions when users request a ride. According to Khosrowshahi, the app correctly predicts the destination in 75% of cases.

"Many people expect these technologies, which are revolutionary — and AI is revolutionary — to have a major impact," Khosrowshahi stated. "Instead, it will be thousands of small improvements that we believe will drive growth in the foreseeable future," he added.

Workforce Reduction and Operational Efficiency

Uber is also transforming its internal operations through AI. The company recently laid off about 10% of its customer service employees, justifying this decision as an opportunity to improve efficiency through AI.

During Wednesday's call, Krishnamurthy emphasized that customer service is "a clear area where we should be able to improve the quality of our interactions with customers while reducing our costs."

The company has also established "agent pods" to explore AI applications in various fields, ranging from finance to human resources.

Advances in Consumer AI Applications

Uber's consumer AI applications continue to progress, although they are not all being deployed simultaneously, Khosrowshahi stated. Uber facilitates "hundreds of thousands" of autonomous vehicle rides each week, which represents less than 0.5% of the 300 million rides taken via the app during the same period, the CEO clarified. In contrast, AI is already widely used by many online search users, he added. Unlike AI chatbots, Uber's physical use of AI through robotaxis requires real-world testing and government approvals.

Uber is also exploring other uses of AI to "enhance the fundamental consumer experience," Khosrowshahi said. He mentioned Cart Assistant, a feature that can create a grocery cart in the Uber app from a recipe or shopping list, and adjust the cart based on customer feedback.

Cart Assistant users have cart sizes that are twice as large as those of users who did not use the AI tool, Khosrowshahi stated. "You can expect AI to contribute to average order size, quality and reliability of our service, as well as presenting the right product at the right time," he concluded.