Uber's AI spending faces scrutiny over lack of ROI

At a glance:

• Uber exhausted its annual AI budget in four months of 2026
• CEO questions connection between token consumption and consumer features
• $3.4 billion R&D spend in 2025, 9% increase from 2024

Uber's AI Investment Struggles

Uber's president and COO, Andrew Macdonald, recently expressed doubts about the return on investment for the company's aggressive AI spending. Speaking at Rapid Response, Macdonald stated that while token consumption for tools like Claude Code has surged, there's no clear link to enhanced consumer features. "The link is not there yet," he emphasized, noting that the company is struggling to justify the financial outlay without tangible product improvements. This disconnect has raised concerns about the strategic value of Uber's AI initiatives, particularly as the company faces mounting pressure to demonstrate measurable outcomes.

The issue isn't just about spending—it's about alignment. Uber allocated $3.4 billion to research and development in 2025, a 9% jump from the previous year. However, Macdonald argued that without a direct correlation between AI investments and user-facing innovations, the trade-off between token costs and human resources becomes increasingly difficult to rationalize. "We’re going to have to start talking about token consumption and the associated cost versus headcount," he said, highlighting a potential shift in how Uber evaluates its tech expenditures.

Financial Pressures and Strategic Shifts

Uber's financial strategy appears to be pivoting as AI costs escalate. The company's CEO, Dara Khosrowshahi, has indicated that increased AI investments are being offset by reduced hiring. This approach reflects a broader trend in tech where firms balance automation with budget constraints. However, Macdonald's remarks suggest skepticism about whether this strategy will yield long-term benefits. "If you’re not able to draw a direct line to useful features," he warned, "that trade becomes harder to justify."

This financial strain isn't unique to Uber. Many tech companies are grappling with the economics of AI, where high upfront costs often outpace immediate returns. Uber's experience could serve as a case study in how organizations navigate the complexities of AI adoption. The company's focus on token consumption metrics may signal a move toward more data-driven decision-making, but the lack of clear outcomes raises questions about the sustainability of such investments.

The Human vs. Token Dilemma

Macdonald's comments underscore a fundamental tension in AI development: the trade-off between automated systems and human labor. By prioritizing token-based AI tools, Uber aims to reduce reliance on human employees, but this approach risks undermining the quality or utility of its offerings. The president's skepticism about the connection between token usage and feature delivery highlights a potential flaw in this strategy. Without concrete evidence that AI investments translate to user value, Uber may face backlash from investors or stakeholders.

This dilemma is not new. Tech companies have long debated the efficiency of AI versus human labor, but the current landscape is more complex. AI tools like Claude Code require significant computational resources, driving up costs. For Uber, the challenge is twofold: ensuring these tools deliver meaningful improvements while managing the financial burden. The company's recent budget exhaustion suggests that the scale of investment may be outpacing its ability to generate proportional returns.

Broader Implications for AI Investment

Uber's struggles with AI spending could have ripple effects across the tech industry. As companies pour resources into AI, the pressure to demonstrate ROI will intensify. Uber's case may prompt other firms to reevaluate their AI strategies, particularly in areas where metrics are hard to quantify. The emphasis on token consumption and headcount trade-offs could become a standard metric for evaluating AI projects.

However, this shift also risks stifling innovation. If companies prioritize cost-cutting over experimentation, they may miss out on breakthroughs that require sustained investment. Uber's experience serves as a cautionary tale about the dangers of overcommitting to AI without a clear path to value. The company's future success may depend on its ability to balance aggressive spending with measurable outcomes.

What's Next for Uber's AI Strategy?

Moving forward, Uber will need to address the disconnect between its AI investments and user benefits. Macdonald's remarks suggest a potential recalibration of priorities, but the company must navigate this carefully. One possibility is a greater focus on niche applications where AI can deliver clear, quantifiable advantages. Alternatively, Uber might reduce its AI spending in the short term to stabilize its budget while exploring more targeted use cases.

The broader tech community will likely watch Uber's next moves closely. If the company can demonstrate a viable path from token consumption to consumer features, it could set a precedent for others. Conversely, continued struggles may lead to a more cautious approach to AI investment across the industry. For now, Uber's experience highlights the challenges of scaling AI in a way that aligns with business objectives.

Conclusion

Uber's AI spending crisis reflects a growing pain point in the tech world. As AI tools become more powerful but also more expensive, companies must find ways to justify their investments. Uber's experience underscores the importance of aligning AI initiatives with clear, measurable outcomes. While the path forward is uncertain, the lessons learned from this case could shape how organizations approach AI in the years to come.