Is Your AI Spend Out of Control? What Uber’s AI Budget Crisis Can Teach Every Business

If your team is using AI tools without spending limits, a formal usage policy, or visibility into costs, you’re at risk of the same problem Uber just ran into. The company burned through its entire annual AI budget in four months and had to scramble to put guardrails in place after the fact. You don’t want to be in that position.

According to a June 2026 report from Bloomberg, Uber instituted a $1,500-per-employee monthly cap on agentic AI coding tools after discovering that costs had spiraled far beyond projections. The company had encouraged staff to use AI “as much as possible,” even ranking internal usage competitively on leaderboards. It worked, maybe too well.

For most businesses, especially those without dedicated IT leadership, the Uber story is a warning sign worth paying attention to now.

The Real Problem Wasn’t AI, It Was a Lack of Governance

Uber’s AI investment didn’t fail because the tools didn’t work. It ran into trouble because there was no framework for managing how those tools were used, by whom, and at what cost. That’s a governance problem, not a technology problem.

The company’s COO acknowledged that it was “very hard to draw a line” between AI usage and actual business results. That’s a significant admission from a global tech giant, and it’s the same challenge businesses of all sizes are facing right now.

A 2026 Bain & Company survey found that AI is delivering less cost reduction than most companies predicted. The ROI is still largely theoretical for many organizations, which means every dollar spent on AI tools must be tied to a measurable outcome.

What “AI Governance” Looks Like in Practice

You don’t need a massive IT department to manage AI responsibly. What you need is a clear set of policies and the right partner to help enforce them. Here’s what that looks like:

• Usage limits by role and department: Not every employee needs access to the same tools at the same tier. Setting role-based access controls prevents runaway spend and keeps AI where it adds real value.
• Cost visibility dashboards: Your organization should have a real-time view of AI tool usage and spend before issues arise, not after.
• Approval workflows for high-usage scenarios: Define when exceptions are allowed and who has the authority to approve them. Unchecked exceptions are where budget blowouts happen.
• Acceptable use policies: What data can employees feed into AI tools? What outputs need human review before use? These questions need written answers, not assumptions.
• Regular ROI reviews: Tie AI usage to outcomes. If a tool isn’t producing measurable results within a defined timeframe, that’s information you need, and you need it before the annual budget cycle, not after.

The Risk of “AI as Much as Possible” Thinking

There’s a dangerous assumption that greater AI use automatically equals greater productivity. Uber’s experience demonstrates that it’s not true. Adoption without accountability creates waste, not output.

For professional services firms, manufacturers, healthcare organizations, and other businesses with real compliance obligations, there are additional risks layered on top of cost. Employees using unapproved AI tools or sharing sensitive data with consumer-grade AI platforms can expose your organization to regulatory and legal liability.

AI governance isn’t a restriction on innovation. It’s what allows you to innovate sustainably, without exposing the business to financial or compliance risk.

Yeo & Yeo Technology: Helping Businesses Implement AI Responsibly

As a managed IT services provider, Yeo & Yeo Technology works with organizations across Michigan to help them implement AI responsibly. That means building the policies, controls, and visibility tools you need before costs become a crisis.

Whether you’re just beginning to deploy AI tools across your team or you’re already seeing usage spike and wondering where the budget is going, we can help you establish a governance framework that makes sense for your size, your industry, and your risk tolerance.

Ready to get your AI strategy under control before it gets away from you?

 Contact Yeo & Yeo Technology to start the conversation.

Frequently Asked Questions

What is AI governance, and why does my business need it?

AI governance is the set of policies, controls, and oversight processes that determine how AI tools are used inside your organization. It covers things like which employees have access to which tools, what data can be shared with AI platforms, how costs are monitored, and how AI-generated outputs are reviewed. Without governance, AI adoption can create compliance gaps, budget overruns, and data security risks, exactly what Uber experienced in 2026.

How much should my business be spending on AI tools per employee?

There’s no universal answer; it depends on the tool, the role, and the measurable output you’re trying to achieve. What matters is that spending is tracked, tied to outcomes, and reviewed regularly. Uber capped its internal AI spend at $1,500 per employee per month per tool after costs spiraled out of control. Most small and mid-sized businesses will operate at far lower thresholds. The key is having a defined number and the visibility to enforce it.

Can employees use personal AI tools like ChatGPT for work tasks?

This is one of the most important questions to address with a written policy. Consumer-grade AI tools may not meet the data privacy or security standards your business requires, especially in regulated industries. Employees using unauthorized tools to process client data, financial records, or protected health information can expose your organization to serious liability. An acceptable use policy, combined with approved tool alternatives, is the right approach.

How do I measure ROI on AI tools?

Start by defining what you’re trying to achieve before you deploy a tool, not after. Are you looking to reduce the time spent on a specific task? Improve response times? Reduce headcount requirements? Quantify the baseline, set a timeframe for evaluation, and track actual outcomes against it. If the numbers aren’t there after a defined period, that’s a signal to adjust your approach, not to keep spending.

What’s the difference between a managed AI strategy and just buying AI subscriptions?

Buying subscriptions gets you access to tools. A managed strategy gets you results. The difference is the policy layer, the governance, cost controls, training, and ongoing review that turns AI access into measurable business value. Yeo & Yeo Technology helps organizations build that layer so AI investments don’t become a budget liability.