The Numbers Are Real: AI Is Erasing Jobs Faster Than Markets Admit

Goldman Sachs released a sobering April 2026 assessment: AI is eliminating roughly 16,000 net jobs per month in the United States, with AI substitution alone responsible for erasing approximately 25,000 positions monthly. This isn’t speculative. Companies like Cognizant and Pinterest are now naming AI explicitly in regulatory filings as the reason for layoffs—“Project Leap” and “reallocating resources to AI-focused teams” have replaced the vague language of “restructuring.”

Four months into 2026, corporate America filed nearly 1,600 layoff announcements affecting over 128,000 workers. While this represents a 5% decline year-over-year, the critical shift is visibility: AI is no longer a background justification. It’s the official reason appearing in SEC filings and press releases.

Early-Career Workers Are Bearing the Weight

The crisis is most acute for workers at the start of their careers. A November 2025 study by Erik Brynjolfsson and Stanford’s Digital Economy Lab found a 16% decline in early-career employment across the most AI-exposed occupations since late 2022. Entry-level hiring at the top 15 tech companies fell 25% from 2023 to 2024—a decline that continued through 2025 and into 2026.

This matters for Ireland and Europe because knowledge work—software development, data analysis, financial services, business process outsourcing—represents a significant portion of employment in Ireland’s multinational tech sector. If early-career hiring is collapsing in Silicon Valley, it’s collapsing here too.

The Labour Market Is Frozen, Not Booming

Economists describe a peculiar phenomenon lasting over a year: a “low hire, low fire” environment. Hiring is suppressed. Firing is suppressed. Quit rates are suppressed. The market isn’t choosing between a hiring boom or recession—it’s frozen by uncertainty about what AI will actually do to headcount.

This uncertainty is paralyzing workforce planning. Companies can’t decide whether to invest in training or reduce headcount, so they do neither. Workers can’t tell whether their skills will be valuable in two years, so many are treading water.

Europe’s Regulatory Framework Doesn’t Address Displacement

The EU AI Act now requires employers to ensure staff have “sufficient AI literacy,” which should theoretically drive training investment. But the Act’s transparency and risk-management requirements—critical for safety—say almost nothing about workforce transition, retraining subsidies, or labour market coordination.

Ireland’s August 2026 AI transparency deadline focuses on detection and disclosure. Europe’s December 2027 and August 2028 compliance crunch prioritises high-risk system governance. Neither addresses the fact that 92 million jobs will be displaced globally by 2030, according to the World Economic Forum, even if 170 million new roles are created.

The Mismatch Problem Is Europe’s Real Challenge

The World Economic Forum projects net job creation of 78 million globally by 2030. But here’s the catch: the jobs lost and the jobs created don’t overlap. A logistics worker displaced by AI-driven warehouse automation won’t automatically become a prompt engineer. An accounts payable clerk won’t pivot to AI safety research without significant retraining and income support.

Europe has stronger social safety nets than the US, but it hasn’t yet coordinated a labour transition strategy at scale. Ireland, with its reliance on foreign direct investment in tech, is particularly exposed: if multinationals reduce early-career hiring, Ireland loses the talent pipeline that has historically moved into management and leadership roles.

What Needs to Happen Now

Europe’s AI regulation must be paired with active labour market policy: sector-specific retraining programmes, income support for displaced workers, and incentives for companies to invest in workforce transition—not just AI compliance.

Ireland’s policymakers should begin mapping which occupations face the highest displacement risk and launch pilot retraining initiatives before August 2026. The regulatory framework is in place. The workforce strategy is not.

Source: Goldman Sachs Economic Research / Stanford Digital Economy Lab