Cloudflare just handed corporate America a new script for layoffs, and it is worth reading closely, because every company doing this is about to copy the wording. When Cloudflare cut 1,100 jobs in May 2026 - roughly 20% of its workforce, and the first mass layoff in its 16-year history - CEO Matthew Prince and President Michelle Zatlyn didn't call it cost-cutting. They called it architecture: "we have to be intentional in how we architect our company for the agentic AI era." That happened in the same quarter Cloudflare posted $639.8 million in revenue, its highest ever, up 34% year over year. The stock still dropped more than 20% anyway. Read that combination again: record revenue, record cuts, framed not as retrenchment but as a redesign.
Amazon is running the same play at a much larger scale. Since October 2025 the company has eliminated close to 30,000 corporate roles - about 14,000 in the initial round, another roughly 16,000 announced this year - hitting AWS professional services, Alexa AI, Prime Video and Studios, and Amazon Pharmacy. And Amazon isn't shrinking to save money either. It's raising capital expenditure from $131.8 billion in 2025 to a planned $200 billion in 2026, almost entirely to build out cloud and AI infrastructure. The pattern at both companies is identical: fewer people, more capital poured into the systems that replace them, and a public relations story that insists this is evolution rather than elimination.
Inc.com's Scott Hutcheson, writing about the wave hitting Oracle, Amazon, Cloudflare, and Block together, put his finger on the actual claim being made: "AI is not simply changing the tasks people perform. It is changing the organizational assumptions on which many companies were built." That is a much bigger statement than "we automated some tickets." It says the org chart itself - how many people you need per function, per team, per layer of management - is being rewritten, and the answer keeps coming back lower. Cloudflare says its internal AI usage grew more than 600% in three months, that its R&D team runs on AI coding tools, and that 100% of deployed code now passes through autonomous AI agent review. Prince describes the productivity gain as going "from a manual to an electric screwdriver," with some workers now doing "2 to 100 times" what they did before.
Here is the part that should bother anyone who has spent a career building expertise instead of a following: none of this is being sold as bad news to shareholders, and it is being marketed as inevitable rather than chosen. Prince has promised Cloudflare will have "more employees than we did at any point in 2026" by 2027 - which sounds reassuring until you notice it says nothing about which employees, doing what kind of work, at what pay, or where. A company can grow headcount while permanently eliminating an entire tier of mid-career technical jobs and replacing them with a smaller number of AI-supervision roles. That's not a hypothetical; it's what "leaner and less complex," Amazon's own language for its restructuring, actually means in practice.
This is where the AI-layoffs story connects to something that's been building for years and rarely gets named in the same breath: the steady replacement of experienced domestic technical talent with cheaper labor, whether that's an offshored team or a visa-dependent one, dressed up as efficiency. AI is a genuinely new lever, but it's being pulled by the same hand that has spent two decades treating American engineering and IT talent as a cost line to be minimized rather than a workforce to invest in. When a company can point to an algorithm instead of a wage arbitrage strategy, the layoff gets a much friendlier press cycle - "AI-driven restructuring" reads a lot better than "we found it cheaper to not employ Americans." The mechanism changes; the outcome for the worker who gets cut doesn't.
None of this means the technology isn't real or that productivity gains are fake - Cloudflare's numbers on code review and deployment speed are specific and plausible. What's worth pushing back on is the framing that lets companies collect record revenue, cut a fifth of their staff, and describe it as something that happened to them rather than something they chose to do, in a labor market that was already brutal for US tech workers before a single AI agent wrote a line of code. If "architecting for the agentic AI era" becomes the industry's preferred euphemism for headcount reduction, workers deserve at least as much clarity about what's actually being optimized for as shareholders got on the earnings call.
Hero image: "Herman Miller Office Cubicles" by heroofficesystems.com, via Wikimedia Commons, licensed CC BY-SA 4.0.
Read the original reporting: Inc.com on AI layoffs at Oracle, Amazon, and Cloudflare, TechCrunch on Cloudflare's record revenue and layoffs, and SecurityWeek's coverage of the Cloudflare restructuring.