Uber president and chief operating officer Andrew Macdonald has voiced growing concerns about the company’s escalating artificial intelligence spending, suggesting that the connection between investment and tangible outcomes remains elusive. In an interview with the Rapid Response podcast, Macdonald stated that despite astronomical growth in certain AI metrics, the link to delivering more useful features for consumers is not yet established.

“That link is not there yet, right? I think maybe implicitly there is more that is getting shipped, but it’s very hard to draw a line between one of those stats and, ‘Okay, now we’re actually producing 25 percent more useful consumer features,’” Macdonald said. He added that while underlying metrics are trending dramatically, the justification for massive AI budgets becomes harder to maintain without visible results.

The comments come after Uber reportedly exhausted its annual AI budget within the first four months of 2026. The company allocated significant resources to research and development, spending $3.4 billion in 2025—a 9 percent increase over the previous year. Earlier this month, Uber CEO Dara Khosrowshahi indicated that the company was offsetting rising AI costs by reducing its human workforce, hiring fewer employees as AI capabilities expanded.

The Trade-Off Between Token Consumption and Headcount

Macdonald emphasized that the real challenge lies in the trade-off between token consumption—a measure of AI usage, particularly in coding assistants like Claude Code—and headcount. “We’re going to have to start talking about token consumption and the associated cost versus headcount,” he said. “So if you’re not actually able to draw a direct line to how much useful features and functionality you’re shipping to your users, that trade becomes harder to justify.”

This perspective reflects a growing recognition among tech leaders that AI spending must be evaluated with the same rigor as any other operational cost. Uber’s experience is not unique but is particularly striking given the company’s history of aggressive investment in autonomous vehicles, dynamic pricing algorithms, and now generative AI tools. The firm uses AI for ride-matching, route optimization, fraud detection, and customer support, yet executives struggle to quantify whether these applications justify the billions spent.

The Broader AI Investment Debate

The skepticism from a major tech executive highlights a broader industry debate. Over the past two years, companies across sectors have poured billions into AI capabilities, often without clear evidence of productivity gains or customer satisfaction improvements. Major cloud providers like Microsoft, Google, and Amazon have reported skyrocketing AI-related revenues, but the actual impact on end users remains mixed. Analysts have begun questioning whether the current spending spree is sustainable, especially given the high cost of training and running large language models.

Uber’s specific concerns revolve around Claude Code, an AI coding assistant developed by Anthropic. The company’s developers rely on the tool to write and debug code, leading to massive increases in token consumption. However, Macdonald noted that translating those tokens into measurable improvements in consumer-facing features has proven difficult. “I think over the coming quarters and years, maybe that will become clearer, but I think today it’s hard, even if some of the underlying metrics are trending in a really astronomical direction,” he said.

Historical Context and Career Highlights

Andrew Macdonald joined Uber in 2014 and has held several leadership roles, including regional general manager for Asia Pacific and head of Rides and Eats. He was promoted to president and COO in 2023. Under his tenure, Uber expanded its delivery business and pursued partnerships in autonomous driving. The company has invested heavily in AI for decades—its core ride-matching algorithm was an early success story—but the explosion of generative AI since 2022 has introduced new costs and complexities.

Uber’s CEO Dara Khosrowshahi has publicly stated that the company is “all in” on AI, but also warned that the technology must deliver tangible business value. In previous earnings calls, he noted that AI could reduce the need for certain roles, such as software engineers and support staff. The tension between innovation and efficiency is now playing out at the highest levels, with Macdonald’s remarks signaling a potential shift in strategy.

Industry Reactions and Comparisons

Other tech companies are facing similar scrutiny. Google’s parent Alphabet has seen its AI-related costs surge, while revenues from cloud services have grown more slowly than anticipated. Microsoft’s massive investment in OpenAI has produced mixed results, with some analysts questioning the return on its multibillion-dollar partnership. Meanwhile, startups like Jasper and Grammarly are struggling to prove that AI-powered features justify premium pricing.

In the transportation sector, rivals like Lyft and Didi are also investing in AI, but with less visibility. Uber’s scale makes its spending particularly notable. The company operates in over 70 countries and processes millions of trips daily. Any improvement in efficiency or customer experience could translate into significant financial gains, but the lack of direct correlation between AI spending and features remains a concern.

Future Outlook

Macdonald expressed hope that the connection between AI spending and feature delivery will become clearer over time. He emphasized the need for better metrics—beyond token consumption—that tie directly to user outcomes. “Maybe implicitly there is more that is getting shipped, but it’s very hard to draw a line,” he noted. Uber has begun experimenting with new ways to measure the impact of AI, including controlled A/B tests and customer satisfaction surveys tied to specific AI-driven changes.

The company is also exploring alternative AI models and more efficient deployment strategies. For example, it is testing smaller, task-specific models that require fewer tokens than general-purpose tools like Claude Code. Additionally, Uber is investing in on-device AI processing to reduce cloud costs and latency. These efforts reflect a broader industry trend toward optimizing AI spending rather than simply increasing it.

Macdonald’s comments come at a time when investor pressure is mounting for tech companies to demonstrate profitability. Uber has only recently achieved consistent quarterly profits, and any sign that AI spending is not delivering returns could impact its stock price. The company’s leadership appears cautious, balancing the need to innovate with the imperative to control costs. As Macdonald put it, “If you’re not actually able to draw a direct line to how much useful features and functionality you’re shipping to your users, that trade becomes harder to justify.”

Source: The Verge News