The warnings about AI-driven job losses are no longer theoretical. A growing body of evidence from Goldman Sachs, the IMF, Anthropic, and major employers suggests that artificial intelligence could eliminate between 10 and 27 million jobs in the United States within the next few years — with entry-level white-collar positions facing the greatest immediate risk.
The Numbers Are Getting Real
Goldman Sachs published one of the most granular analyses of AI's employment impact to date, finding that AI substitution is currently wiping out roughly 25,000 jobs per month in the United States. AI augmentation — where the technology makes existing workers more productive — adds back about 9,000 positions. That leaves a net loss of approximately 16,000 jobs per month, or nearly 200,000 positions per year.
The analysis separated two competing effects: substitution, where AI replaces human workers outright in roles like insurance claims processing and bill collection, and augmentation, where AI handles some tasks but human judgment remains essential, as in law, medicine, and construction management.
Anthropic CEO Dario Amodei has been even more direct. He told reporters that AI could wipe out half of all entry-level white-collar jobs and push unemployment to Depression-era levels of 10 to 20 percent within the next one to five years. Based on current workforce numbers, that translates to between 10 million and 27 million jobs.
Block's 40% Cut Sets the Tone
The scale of potential displacement was thrown into sharp relief when Block, the payments company led by Jack Dorsey, laid off 4,000 employees — roughly 40 percent of its workforce. Dorsey was blunt about the reason: AI tools paired with smaller, flatter teams are enabling a fundamentally new way of working.
If that ratio were applied across the S&P 500, which collectively employs around 29 million people, the result would be 12 million job cuts. That scenario is extreme, but it illustrates the scale of what becomes possible when companies decide that AI can handle functions previously performed by large teams.
Amazon has moved in a similar direction, announcing 30,000 corporate job cuts across multiple rounds. CEO Andy Jassy explicitly cited AI adoption as the driver, while the company simultaneously projects $200 billion in capital expenditure for 2026 — much of it directed at AI infrastructure.
Gen Z Is Taking the Brunt
The impact is falling disproportionately on younger workers. Entry-level hiring at the top 15 tech companies dropped 25 percent between 2023 and 2024, with declines continuing through 2025 and into 2026. The reason is straightforward: AI tools now handle the tasks that companies used to assign to junior employees — writing boilerplate code, drafting emails, creating basic reports, and summarizing documents.
Anthropic's own labor market research found that fewer young people are starting jobs in highly AI-exposed occupations. The job-finding rate for these roles fell by about 14 percent compared with 2022. Meanwhile, 64 percent of Gen Z workers say they are worried about losing their job to AI, compared with 45 percent of millennials and 29 percent of boomers.
Meta and the AI Workforce Paradox
Meta exemplifies the paradox at the heart of AI-driven employment. The company has projected $115 to $135 billion in capital expenditure for 2026 as AI spending surges, while simultaneously restructuring into smaller pods where AI-generated reporting systems replace multiple layers of middle management. Operating margins have fallen from 48 to 41 percent as infrastructure costs climb.
The pattern is consistent across the tech sector. Companies are spending more on AI while employing fewer people — a combination that has investors questioning whether the productivity gains can offset the staggering infrastructure costs before patience runs out.
The Reskilling Gap
The World Economic Forum projects that AI and related technologies will create 170 million new jobs globally by 2030 while displacing 92 million — a net positive of 78 million. But the problem is not the total number. It is the mismatch between the jobs being destroyed and the jobs being created.
The skills gap between a displaced administrative assistant and an AI engineer cannot be bridged by a short training course. While 53 percent of organizations say they prioritize reskilling, only 21 percent believe they are doing it effectively. Over 40 percent of workers will require significant upskilling by 2030, according to WEF projections.
The IMF's managing director described AI as hitting the labor market like a tsunami, warning that most countries and most businesses are simply not prepared for what is coming.
What Comes Next
The debate over AI and jobs is shifting from prediction to observation. The layoffs are happening. The entry-level positions are disappearing. The net job losses are measurable on a monthly basis. Whether the promised new jobs will materialize fast enough to absorb the displaced workers remains the most consequential economic question of the decade.
For now, the data points in one direction: AI is not just changing how work gets done. It is changing how many people are needed to do it.
