Why Microsoft and Uber Hitting an AI Budget Wall Is Great News for Canadian Workers

For months, the narrative has been impossible to ignore: Artificial intelligence (AI) is coming for your job, and it’s only a matter of time until you’re out of work. But two major corporate stumbles this spring tell a more complicated story — and for Canadian workers anxious about their future, it offers something unexpected: a reason to breathe a little easier.

Microsoft — the company that has poured about US$13 billion (C$17.8 billion) into OpenAI and writes up to 30% of its own code using generative AI — reportedly told engineers in a major division to stop using an AI coding tool because the bills got too big. Meanwhile, Uber’s chief technology officer said the company burned through its 2026 budget for Claude Code and Cursor, two leading AI coding tools, in only four months.

The uncomfortable truth is that the technology reshaping the global economy is, right now, too expensive to do what its biggest boosters have promised. For workers — including the millions of Canadians whose jobs may be in the crosshairs — that economic reality is quietly buying time.

What Microsoft Actually Did, and What It Didn’t

In late 2025, Microsoft gave thousands of its people — engineers, product managers, designers and even staff in non-technical roles — access to Claude Code, Anthropic’s command-line AI coding agent. The idea was to let them experiment. It spread quickly, well beyond technical teams.

Then the bills arrived.

Microsoft began cancelling Claude Code licenses across its Experiences and Devices group — the team behind Windows, Microsoft 365, Outlook, Teams and Surface — with a June 30, 2026 cutoff, the last day of the company’s fiscal year. Microsoft then moved its engineers to GitHub Copilot CLI, its more affordable in-house tool.

To be clear: This isn’t Microsoft walking away from AI. Claude models still run inside Copilot CLI. In fact, its broader deal with Anthropic — including a US$5 billion investment (C$6.8 billion) and Anthropic’s US$30 billion (C$41 billion) commitment to buy Azure computing capacity — remains intact, according to Forbes.

The problem is the pricing model. Token-based billing charges per unit of output, and when engineers use an AI agent for hours on complex coding tasks, those tokens — and the associated costs — pile up fast.