AI Powers Record Tech Revenue While Entry-Level Jobs Dry Up

Amazon, Microsoft, Alphabet, and Meta collectively committed $700 billion to AI infrastructure in 2026, nearly double their 2025 capital spending. Over the same months, the U.S. tech sector absorbed more than 142,000 layoffs. Microsoft’s AI business runs at a $37 billion annual revenue run rate, up 123% year-over-year; Google Cloud grew 63% in Q1 to $20 billion. The 2026 Stanford AI Index documents the employment cost: software developer jobs for workers aged 22 to 25 fell nearly 20% since 2024, the first white-collar job category to show documented AI-driven contraction.

The $700 Billion Redirect

Each of the four hyperscalers raised its capital expenditure forecast on Q1 2026 earnings calls. Amazon held at $200 billion. Meta moved from $72.2 billion in 2025 to between $115 billion and $135 billion for 2026, with CFO Susan Li citing infrastructure for Meta Superintelligence Labs and core business as the primary drivers. NVIDIA’s 2026 State of AI survey of more than 3,200 executives found that 86% of respondents plan to increase their AI budgets this year, as enterprise experiments from 2025 convert to full production deployments across legal, financial, and administrative functions.

• $700 billion: combined 2026 capex from Amazon, Microsoft, Alphabet, and Meta, nearly double 2025 levels
• 142,000+: U.S. tech workers laid off industry-wide through late May 2026
• 33%: pace of 2026 tech cuts above the same period in 2025, per Challenger, Gray & Christmas

In 2025, U.S. companies announced more than 1.17 million layoffs, the highest annual total since the 2020 pandemic, with nearly 55,000 directly attributed to AI in official announcements, per Challenger, Gray & Christmas. The 2026 acceleration, running 33% above last year’s pace through April, suggests the industry isn’t moving toward a stable equilibrium.

Challenger, Gray & Christmas tracked 85,411 tech job cuts in the first four months of 2026. Layoffs.fyi counted more than 113,000 tech workers laid off through mid-May across 179 companies, running at about 825 per day. About 48% of those cuts were explicitly attributed to AI by the announcing companies, though economists note that AI has also become a convenient framing for restructuring that would’ve happened regardless of the technology.

The pattern is clearest where specific operational links are stated. Oracle cut 10,000 employees in early April, with analysts expecting the total to reach roughly 20% of its global workforce, immediately after reporting strong quarterly earnings, then announced expanded AI data center investment. CEO Mark Zuckerberg described the productivity equation on Meta’s January 2026 earnings call.

We’re starting to see projects that used to require big teams now be accomplished by a single very talented person.

Zuckerberg, Meta’s co-founder and chief executive, said this while simultaneously projecting the company’s largest-ever annual infrastructure spend.

Junior Developers Take the First Hit

What Stanford Measured

The Stanford Institute for Human-Centered AI’s 2026 AI Index economy chapter, published April 13, produced the labor market data with the most sector-specific precision yet. Employment for software developers aged 22 to 25 fell nearly 20% since 2024, Stanford HAI found, making entry-level software engineering the first white-collar job category to show measurable contraction directly attributable to AI.

Developers aged 30 and older at the same companies saw headcount grow during the same period. One-third of organizations surveyed by Stanford HAI expect AI to reduce their overall workforce in the coming year, with anticipated cuts highest in service operations, supply chain, and software engineering. AI skills now appear in 2.5% of all U.S. job postings, up 55% year-over-year, according to Lightcast labor data compiled for the Stanford report.

Workers with AI skills command an average 56% wage premium over those without, per PwC’s 2025 Global AI Jobs Barometer. That premium has grown as AI tools have proliferated, because demand for workers who can build and manage AI systems is outpacing supply. Goldman Sachs Research separately found that unemployment among workers aged 20 to 30 in tech-exposed occupations has risen by nearly three percentage points since early 2025, notably higher than for older workers in the same occupational groups.

The Work That Vanished

The age concentration isn’t coincidental. AI tools have become reliably capable at exactly the tasks junior engineers are typically hired to do in their first two to three years:

• Writing boilerplate code and basic CRUD (create, read, update, delete) operations
• Scripted testing and routine bug triage
• Straightforward feature implementations against defined specifications
• Document parsing, data formatting, and basic processing pipelines

Senior engineers at large companies are now using AI to handle those tasks directly, without delegating to junior staff. The entry point to software engineering, historically built around low-stakes work that took two to three years to graduate past, is compressing at exactly the moment demand for AI-native skills is accelerating. That’s the gap the 2026 employment data is measuring.

Record Revenue, Shrinking Headcounts

What distinguishes the 2026 layoff wave from prior tech downturns is the financial context. Companies executing the largest cuts are doing so from positions of strength. Oracle reported strong quarterly earnings before announcing cuts of up to 30,000. Meta’s Q4 2025 revenue was a quarterly record before the company cut 8,000 staff.

Salesforce CEO Marc Benioff described the operational shift plainly: AI agents now handle approximately half of the company’s customer interactions, enabling cuts to 5,000 customer support roles. At Block, CEO Jack Dorsey reduced headcount from 10,000 to under 6,000, attributing the decision directly to AI in a shareholder letter, noting that AI intelligence tools had changed what it means to build and run a company. Citigroup has targeted a 20,000-employee reduction as AI-enabled automation absorbs middle-office functions.

Intuit announced 3,000 additional cuts on May 20, representing 17% of its global headcount, the same day Meta began notifying its 8,000 affected employees. California Governor Gavin Newsom signed an executive order the following day directing state agencies to study AI-driven displacement and develop policy recommendations for affected workers.

A Market of Two Speeds

PwC’s 2026 AI Performance Study, surveying 1,217 senior executives across 25 sectors, found that 74% of AI’s economic value is being captured by just 20% of organizations. That group deploys autonomous systems to pursue growth through industry convergence. Most companies are still running AI pilots without measurable financial returns.

For the 80% not capturing significant AI returns, the gap is as much organizational as technological. PwC’s analysis found that companies with the best AI-driven outcomes increase autonomous decision-making at nearly three times the rate of peers and are roughly twice as likely to deploy AI in advanced configurations, either executing multiple tasks within guardrails or operating in fully self-optimizing modes.

Revenue per employee grew 27% in AI-exposed industries versus 9% in the least-exposed sectors, per PwC’s analysis. The differential funds further investment inside winning firms and widens the gap to those still catching up.

Goldman Sachs Research tracked data center construction adding more than 216,000 positions since 2022, running at roughly 9,000 new roles per month by mid-2026 as electrical contractors, HVAC specialists, and commercial construction workers build the physical infrastructure AI compute requires. The buildout is estimated to require approximately 500,000 new U.S. power sector and infrastructure roles by 2030, per Goldman. Well-paid and growing, but they aren’t the destination for a marketing analyst or a junior software engineer whose role was automated.

A Fortune survey of more than 350 public-company CEOs and investors managing $19 trillion in assets found that 66% plan to freeze or cut hiring through the rest of 2026. Of those, 84% acknowledged that meaningful AI return on investment is a multiyear project, suggesting the hiring pause isn’t a short-term correction.

Why This Wave Cuts Deeper

Every major technology wave of the past century destroyed jobs and, over time, created more. Mechanical looms displaced handweavers, and factory floors absorbed them. Mainframes eliminated ledger clerks; a software industry grew in their place. The internet’s arrival wiped out travel agencies as digital marketing expanded. In each case, automation targeted physical or routine tasks, with the cognitive middle absorbing the displaced workers.

Goldman Sachs Research, in its 2026 labor market analysis led by economist Joseph Briggs, updated its estimate of AI task automation exposure from 25% of U.S. work hours in 2023 to 34% in 2026. The categories with the highest exposure are marketing, customer service, document review, graphic design, and entry-level software engineering, the roles where workers displaced from prior automation rounds historically landed.

Goldman’s AI Adoption Tracker estimated gross AI substitution eliminated approximately 25,500 U.S. positions per month in April 2026, while AI-driven job creation added back roughly 9,500. The net figure of approximately 16,000 per month revised downward to around 11,000 by early June as data center construction absorbed additional workers, though Goldman economists noted construction employment is time-bound while AI tool deployment is continuous.

Morgan Stanley Research characterized AI’s labor effect as ‘double-edged’: the same tools that substitute for some workers can amplify the output of those who remain. Evidence for augmentation exists in healthcare and engineering, where AI expands what individual specialists accomplish. In software engineering, the 2026 data runs the other way.

Briggs’ base case projects a manageable 0.6 percentage point increase in the unemployment rate if adoption runs over a 10-year window. His published warning is the one the 2026 data is already testing: “If it’s more frontloaded, the impacts on the economy are much larger.” The Stanford HAI data on developer employment suggests the frontloading has started, at least in the first white-collar segment to absorb measurable contraction.

As of early June 2026, 24% of Russell 3000 companies mentioned AI and labor together on their Q1 earnings calls, per Goldman Sachs’ tracker, up sharply from prior quarters. The federal government has issued no workforce mandate comparable to California Governor Gavin Newsom’s May 21 executive order directing state agencies to study AI-driven displacement and develop policy recommendations for affected workers.