Uber Exhausts 2026 AI Budget Four Months In, Questions ROI on Rising Token Consumption

Uber has burned through its entire 2026 artificial intelligence budget just four months into the fiscal year, according to company president Andrew Macdonald, who said in a Saturday interview with Rapid Response that the ride-hailing giant is struggling to connect rising token consumption for Claude Code with measurable consumer benefits.

The budget depletion comes as Uber continues aggressive AI investment following a $3.4 billion research and development spend in 2025 — a 9 percent increase from the previous year. CTO Praveen Neppalli Naga had previously indicated to The Information in April that the company's Claude Code budget specifically was already exhausted for 2026.

Executive Concerns Over AI ROI

Macdonald's comments signal mounting internal pressure over AI return on investment at the transportation company. The president and COO told Rapid Response that despite substantial token consumption increases, Uber has not identified a clear correlation between higher Claude Code usage and more valuable features reaching end users.

The disconnect between AI spending and tangible outcomes reflects broader industry challenges in translating large language model capabilities into consumer-facing value. While token consumption provides a direct measure of AI system usage, the gap between computational intensity and feature utility has become a focal point for executives managing multi-billion-dollar AI budgets.

Workforce Adjustments to Fund AI Push

CEO Dara Khosrowshahi has indicated the company is offsetting increased AI investments through reduced human hiring. This strategic trade-off positions Uber alongside other technology companies prioritizing AI capability development over traditional workforce expansion, though the effectiveness of this approach remains under scrutiny given the ROI concerns raised by Macdonald.

The workforce reduction strategy suggests Uber views AI investments as potentially replacing rather than augmenting human capabilities in certain operational areas. However, the current budget exhaustion and questioned returns complicate this calculation, particularly if token costs continue rising without proportional feature improvements.

Looking at the trajectory here, we have seen this pattern before with previous technology adoption cycles — the initial investment surge often precedes a rationalization phase where companies reassess spending efficiency and demand clearer metrics on capability translation to user value. The internet buildout of the late 1990s and early cloud migrations both followed similar arcs of heavy upfront investment followed by more disciplined ROI analysis.

Autonomous Vehicle Timeline Pressure

Beyond immediate AI spending concerns, Macdonald has described autonomous vehicle development as existential for Uber's long-term viability. In previous interviews, he projected full autonomy deployment within a timeframe spanning "a couple of years to a couple of decades" — a wide range that underscores both the strategic importance and technical uncertainty surrounding self-driving capabilities.

The autonomous vehicle push adds complexity to Uber's AI budget calculations, as both consumer-facing AI features and foundational AV research compete for the same computational and financial resources. The current budget exhaustion may force more precise allocation decisions between immediate feature development and longer-term autonomy investments.

Business Diversification Efforts

Separately, Uber continues expanding beyond transportation into hotels and hospitality services, diversifying revenue streams while AI investments strain existing budgets. This expansion strategy provides potential buffer against transportation market volatility but also requires additional technology infrastructure investment at a time when AI spending has already exceeded projections.

The hospitality expansion represents a hedge against autonomous vehicle disruption — if self-driving technology eventually commoditizes ride-hailing, Uber's platform relationships and logistics expertise could transfer to accommodation booking and management services.

Industry Context and Forward Outlook

Uber's AI budget challenges arrive amid broader enterprise struggles to quantify large language model ROI. Token pricing models create direct cost correlation with usage volume, but measuring output quality and user value remains complex across applications from code generation to customer service automation.

The company's experience with Claude Code specifically highlights the difficulty in translating developer productivity tools into consumer-facing improvements. While these systems can accelerate internal development processes, the path from faster code generation to better rider or driver experiences involves multiple intermediate steps where value can be lost or diluted.

The budget exhaustion four months into 2026 suggests either significant underestimation of AI usage scaling or rapid feature development pace that exceeded initial projections. Either scenario indicates growing pains as enterprises integrate AI capabilities at scale while managing cost containment and measurable outcome delivery.

For Uber, the immediate challenge involves recalibrating AI spending to sustain operations through the remainder of 2026 while maintaining development momentum on both consumer features and autonomous vehicle research. The company's ability to demonstrate clear ROI from existing AI investments will likely determine budget allocation strategies for 2027 and beyond.