Uber’s business model is one of the most AI-driven in Silicon Valley. AI determines ride prices and optimizes routes, among other predictive capabilities. But even with these advanced features, Uber executives are warning about the ride-hailing company’s AI spending.
In a recent interview, quick response Andrew McDonald, Uber’s president and chief operating officer, said on a podcast that it’s difficult to draw a connection between the company’s increased use of Claude Code and innovations aimed at serving consumers.
“We don’t have that connection yet,” he said. “Maybe implicitly we’re shipping more stuff, but it’s very hard to draw the line between one of those stats and ‘Okay, now we’re actually producing like 25% more useful consumer features.'”
The comments follow reports that the company had already used its 2026 AI coding tools budget in just four months after encouraging employees to adopt the technology through internal leaderboards that ranked teams by total AI tool usage. This is the latest development in the complex issues that arise in enterprise AI implementation. Increasing AI usage comes with increased costs, even if the price per AI unit decreases.
“If you can’t actually draw a direct line on how to do it, [many] “When you provide a user with a useful feature, it becomes difficult to justify the transaction,” McDonald said.
Uber isn’t the only company facing this problem. Microsoft reportedly began canceling most of Claude Code’s direct licenses earlier this month. The VergeInstead, engineers now use the GitHub Copilot CLI. Many other business leaders have reversed their initial bullish AI views. Last year, Duolingo CEO Louis von Ahn reversed his views on AI, saying he doesn’t believe the technology will replace the work done by employees.
Uber did not immediately respond. luckThis is a comment request from .
Can companies justify spending on AI?
During an earnings call earlier this month, Uber CEO Dara Khosrowshahi said that about 10% of the company’s committed code is built by autonomous agents. However, he added that the company’s use of AI goes beyond the scope of software engineers.
“Whether it’s our legal team, our marketing team or our developers, these tools are increasingly being used,” he said. “We think it creates employees with superpowers.”
However, increased use of AI also increases costs. According to a recent study by research firm Gartner, the cost for AI companies to infer highly sophisticated AI models will be 90% lower by 2030 than it was in 2025. However, the study also found that agent models require far more tokens per task than standard models, and cheaper tokens do not translate into cheaper enterprise AI because AI providers do not fully pass on lower costs to consumers.
Some AI companies are changing their pricing plans to capture the growth in AI usage. Anthropic has changed its pricing model, moving from a flat fee to a usage-based model. This means that autonomous agents are now charged per token for compute usage. In March, OpenAI CEO Sam Altman articulated the broader direction of the industry in an interview.
“We are seeing a future where intelligence becomes a utility like electricity or water, and people buy it by the meter,” he said.
Another Gartner study predicts that spending on AI agent software will reach nearly $207 billion in 2026, an increase of more than 139% from $86.4 billion in 2025.
Uber spent 3.4% on research and development in 2025, an increase of 9% from 2024. The company spent $951 million on research and development in the first quarter of 2026 alone, an increase of nearly 17% year over year.
Still, Uber isn’t shying away from innovation. McDonald said the company is fully committed to autonomous driving and expects it to become the norm within decades.
“I don’t think my daughters, who are young now, will eventually get a driver’s license,” he said.
