Cloudflare is a company that most people have never heard of, even though it quietly sits behind a large portion of the world's internet traffic, protecting websites from attacks and keeping them fast. On Thursday, it cut 1,100 people, about one in five of its employees, and made its reasoning unusually blunt: AI had made those roles unnecessary. Revenue was up 34% on the same quarter a year ago. The business had never done better.
What makes this different from a normal round of cost-cutting is the honesty. For years, companies used softer language. They talked about "right-sizing," "strategic restructuring," or cleaning up pandemic-era overhiring. That language is now gone. Cloudflare's CEO said plainly that some roles are not the roles a company needs going forward. Other CEOs have said similar things in recent months. The era of polite euphemisms for AI-driven job cuts appears to be over.
The specific roles Cloudflare targeted tell you something important. Salespeople with revenue targets were explicitly protected. Everyone behind them was not. The CEO described it as the "support people" behind frontline staff. That category covers a wide territory: coordinators, administrators, analysts, junior managers, HR generalists, finance support staff, marketing operations people. These are not niche tech jobs. They are the backbone of how almost every mid-sized organisation runs.
A separate but equally important story is happening at the very largest companies. Amazon, Meta, Microsoft, and Google are collectively planning to spend around $725 billion on AI infrastructure in 2026, up 77% from last year. At the same time, they are cutting tens of thousands of workers. The math matters here. Even if Meta fired its entire global workforce tomorrow, the saving would be roughly $27 billion. Its AI infrastructure budget alone is over $140 billion. The layoffs are not paying for the AI spending. They are something else: a deliberate signal about what kind of company each of these firms wants to be.
Citigroup has flagged plans to reduce its headcount by around 20,000 through AI-enabled automation of middle-office functions. Logistics company C.H. Robinson cut 1,400 jobs after rolling out AI tools for pricing and shipment tracking. A payroll software company called Paycom let go of 500 staff and told them directly their roles had been replaced by AI systems. These are not tech companies in any traditional sense. The pattern is moving across industries.
The honest version of what is happening has two parts, and both are true at the same time. First, AI tools genuinely are making certain categories of work faster and cheaper to perform with fewer people. The productivity gains Cloudflare described, with internal AI use up sixfold in three months, are real and measurable. Second, some of what is being branded as an AI transformation is also financial discipline dressed in new language. Companies that overhired during the pandemic boom now have a culturally acceptable way to cut, and they are using it.
What this means practically is that the roles most exposed are not the glamorous ones. They are the process-heavy, coordination-heavy, and reporting-heavy functions that make organisations run day to day. Junior analysts, back-office coordinators, first-line HR support, marketing operations, financial reporting staff. These are the functions where AI tools have become genuinely capable over the past 18 months, and they are where the cuts are landing.
Cloudflare's stock fell 18% the day of the announcement, despite the strong results. Investors were not celebrating. That reaction is worth noting. The market is not yet convinced that cutting 20% of your people while revenue grows is a straightforward win. The productivity gains are real, but so is the execution risk of rebuilding an organisation around tools that are still evolving fast.