Only 46 Out of 25,000 NY Layoffs Blamed on AI-Rest Is Washing

Only 46 out of roughly 25,000 laid-off workers in New York State — two-tenths of a percent — were flagged as affected by AI in mandatory WARN filings over the past year. New York added an AI disclosure checkbox to WARN Act filings in March 2025. In the full first year, more than 160 companies filed notices, and not a single one checked the box. Nespresso eventually did, accounting for those 46 workers, but that’s it. The state Department of Labor confirmed the number. If AI were really replacing software engineers at scale, you would expect at least a few tech companies to admit it under oath.

The Headlines vs. The Reality

Block CEO Jack Dorsey blamed AI for laying off 4,000 employees in February, claiming the technology enabled “smaller and flatter teams.” A data scientist on the Cash App team reported “very limited gains in productivity” and refused a 75% retention raise. Snap’s Evan Spiegel told employees AI generated 65% of new code while cutting 1,000 jobs — yet the cuts hit the augmented reality division, not engineering. Intuit axed 3,000 roles alongside deals with Anthropic and OpenAI, but the CEO explicitly said “none of it had to do with AI” — the cuts targeted coordination-heavy management layers. Every high-profile AI-driven layoff story examined by Arvind Narayanan and Sayash Kapoor collapsed under scrutiny.

The 10x Gap Between Anticipation and Execution

A Harvard Business Review survey of over 1,000 global executives found that 21% had made large headcount reductions “in anticipation of” AI, while another 39% made low or moderate anticipatory cuts. Only 2% had already made large reductions tied to actual AI implementation. That 10x gap — anticipatory vs. actual — tells you executives are succumbing to the same AI-replacement narrative they’re selling. Forrester principal analyst J. P. Gownder puts it plainly: when you ask companies preparing AI-driven layoffs if they have a mature app ready to fill those jobs, nine out of ten times the answer is no. Even the hiring managers are in on the con — 59% admitted they emphasize AI when explaining layoffs because it plays better with stakeholders than citing financial constraints.

Why the Middle of the Sandwich Isn’t the Whole Meal

Narayanan and Kapoor frame software development as a “decide-execute-deliver sandwich.” AI compresses the execution layer — the meat in the middle — but the deciding and delivering layers resist automation in ways that capability improvements alone won't overcome. You can prototype a feature in minutes with a coding agent; turning it into a finished, deployed, maintained product still requires human judgment, coordination, and ownership. That’s why demand for software engineers isn’t cratering even as AI agents get better at writing code. The next essay in the series will examine why individual careers might still get rocky, but the broad picture is clear: mass AI displacement of engineers is a story that the data refuses to tell.

Source: Why AI hasn't replaced software engineers, and won't
Domain: normaltech.ai