The AI That Wasn’t There
Companies keep blaming artificial intelligence for layoffs. The real story is messier — and states are starting to notice.
A new report this week from a major recruiter depicts layoffs across twelve states, and paints a portrait of an economy in structural distress. And the explanations pouring out of corporate communications departments tend to lean one way: AI is transforming our business, we’re restructuring for the future, people will be affected. The formula has become the quintessential public-relations choreography of 2026.
But a closer look at the state-by-state data reveals something more complicated than robots eating jobs. What we’re actually seeing is AI-washing on a national scale — the practice of attributing layoffs to technological transformation when the real drivers are older, more mundane, and harder to spin: offshoring, revenue decline, patent cliffs, tariff pressure, and federal budget cuts creating downstream shocks across state economies.
On the one hand, it may come as a relief to think that AI isn’t actually able to replace human workers en masse quite yet. On the other hand, it’s alarming just how comfortable employers are becoming in citing AI when they do slash jobs.
In Washington State, the AI threat looks pretty real. Washington recorded 78,943 announced job cuts in 2025 and posted its first annual employment decrease in the Seattle area since 2009. Amazon cut 14,000 positions in October 2025, then another 16,000 in January 2026 — citing, in part, AI’s “transformative” potential. Microsoft eliminated 12,000 across two rounds. (Boeing, whose crisis has nothing to do with artificial intelligence, is cutting 17,000 as it absorbs a machinist strike, a 737 Max production freeze, and a $6 billion quarterly loss.) Meanwhile, Washington State has been developing rules that would require companies to disclose when AI affects employment decisions. As I’ve reported previously for Hard Reset, the companies most aggressively lobbying against those requirements have been conducting layoffs at the same time.
In some states, job losses are just structural. New Jersey posted 66,923 announced job cuts in 2025, with unemployment reaching 5.4%. The pharma sector has been hit especially hard: Bristol Myers Squibb filed five separate WARN notices totaling 1,156 layoffs. Novartis cut 485. Johnson & Johnson eliminated 231. Novo Nordisk laid off 265. Combined pharma layoffs in the state exceeded 2,200 in 2025 alone. The employers have spoken publicly of a push for efficiency and productivity. But the real problem for those companies is a patent cliff — a wave of blockbuster drugs losing exclusivity simultaneously, forcing companies to slash overhead. That’s an accounting story. In New York, where 109,030 job cuts were announced in 2025 and Wall Street profits simultaneously nearly doubled to $49.6 billion, financial services layoffs outpaced hiring for 19 consecutive months while the sector posted record earnings. Again, that’s not AI disruption. That’s profit optimization.
But in other states, layoffs that turn out to be about accounting and competition are often dressed up in technological terminology. Take Lowe’s, headquartered in North Carolina. On February 13, the company eliminated roughly 600 technology jobs — product managers, designers, QA analysts, store operations staff — via a pre-recorded call attached to a calendar invite where all attendees were marked “Optional.” Some employees discovered they’d been laid off when their laptops stopped connecting. CEO Marvin Ellison has said publicly that “AI isn’t going to fix a hole in your roof,” hinting that corporate tech workers were in his sights. But according to great reporting from The McClure Standard, the teams actually cut had been building the software that store associates use every day. And it turns out Lowe’s Global Capability Center in Bengaluru, India, which employed about 1,000 people when its current CIO arrived in 2018, now employs more than 4,700. That offshoring operation was scaling in 2019, 2020, and 2021, long before any executive had a chatbot to point to. As one laid-off employee told The McClure Standard: “I think AI and RTO [return to office] are the cover for the business expenditure angle.”
Still, North Carolina is bracing for the real thing. The North Carolina Technology Association’s 2026 outlook projects that 61% of the state’s tech workforce — about 286,000 jobs — could be heavily affected by AI adoption. Software developers, data scientists, and research analysts top the exposure list. The NCTA’s report carefully distinguishes between AI exposure and AI replacement, but considering that software development alone employed 54,700 people in 2024, “these are roles that have historically driven growth in the state’s tech sector,’’ the report says. “If exposure leads to even modest levels of job replacement, it could significantly impact the state’s employment figures.” And when genuine displacement does arrive in North Carolina, a lack of state regulations there mean workers will have no established right to know whether an algorithm drove the decision.
The displacement that states should be preparing for is real. But in the meantime, as companies continue to woo shareholders with talk of AI doing the work, we’ll need to sharpen our ability to read between the lines. And sprinting too fast into job protections that only center on AI could serve to make AI-washing even more effective. Retraining programs built around AI displacement won’t help someone laid off because their employer moved the job overseas. Disclosure laws targeting automated hiring tools won’t help a pharma researcher terminated because a patent expired.
The companies have every incentive to keep the narratives blurred. The workers and the policymakers trying to respond on their behalf will have to move quickly to bring the threats to jobs into focus.