On March 5, 2026, Bloomberg reported that Oracle was planning to cut thousands of jobs across multiple divisions.

Within two days, a familiar narrative took hold across global tech media:

Estimates suggested as many as 30,000 positions could disappear — up to 18% of the company's workforce — freeing billions for data-centre investment.

The implication was clear. Artificial intelligence is eliminating jobs.

But a closer reading of the underlying financial reality tells a different story.

Oracle is not replacing tens of thousands of employees with software agents. It is managing a balance sheet under strain. The company has accumulated more than $100bn in debt after an aggressive borrowing spree to finance infrastructure projects for major AI customers. It has already disclosed a $1.6bn restructuring charge — largely severance costs — and is reportedly considering asset sales.

Oracle's situation is not exceptional. It reflects a broader restructuring template now visible across large technology firms.

Headline layoff figures rarely correspond to the number of employees actually dismissed. Announcements often combine multiple adjustments: cancelled hiring requisitions, denied backfills, internal redeployments and shifts to contract labour. When Meta announced its wave layoffs in 2023, the plan included eliminating 10,000 roles and closing 5,000 open positions. The "headline layoff count" combined actual exits with jobs that simply never got posted.

Oracle's own FY2025 10-K discloses that 29% of open non-entry-level positions were filled internally — roles disappearing on one side of the org chart while people reappear on the other. The same filing references multiple restructuring plans with staggered accruals and cash payments across quarters. Not a single day of layoffs. A rolling portfolio change in labour demand.

The effect is organisational churn rather than simple contraction.

Annual filings illustrate the scale of this dynamic. I pulled the 10-K employee disclosures for the five companies that have dominated layoff headlines over the past two years.

Every single one of these companies grew headcount during the period in which layoffs dominated the news cycle.

Amazon's case is particularly instructive. In 2020 and 2021, the company hired 810,000 people in two years, more than doubling in size. What we are watching now is not AI-driven contraction. It is the cleanup of the most chaotic hiring binge in corporate history.

Large numbers of departures coexist with continued hiring — often in different geographies, seniority levels or business units.

The persistence of layoff headlines reflects a powerful financial incentive structure — one that is worth understanding in detail, because it explains why the pattern repeats.

Payroll is typically the largest operating expense in technology companies. When restructuring is announced, cost reductions flow quickly into operating margins and earnings per share — often within 90 days. Salaries, benefits, equity compensation, office costs: all removed from OpEx within the quarter the cuts take effect.

The restructuring itself is reported as a one-time charge. Analysts and investors routinely strip these charges out of their valuation models. They look at "adjusted earnings," not GAAP earnings.

Oracle's case illustrates the mechanics precisely:

Severance is noise. Margin improvement is signal. Wall Street ignores the first and rewards the second.

If the layoffs are framed as "AI-driven efficiency," the narrative becomes doubly rewarding: not only is the company more disciplined on costs, it is also perceived as forward-looking on technology. The stock receives a discipline bump and an innovation premium simultaneously.

The timing reinforces the incentive. Reuters reported Amazon's 30,000-job cut plan ahead of an upcoming earnings report. Oracle's "thousands" of planned cuts surfaced in the same reporting cycle as its quarterly results. Meta's January 2025 cuts landed just before earnings season.

Corporate messaging increasingly attributes workforce reductions to artificial intelligence. Yet empirical research paints a more nuanced picture.

A 2025 survey by the Federal Reserve Bank of New York found that among service firms adopting AI tools:

*The reduction does not target the most qualified roles. It targets the roles that graduates used as their first entry point into the workforce.

A February 2025 NBER working paper (Hampole et al.) explains why the gap between narrative and reality is so wide: AI does substitute for specific tasks — the data confirms this — but productivity and growth effects tend to offset those losses elsewhere. The net impact on employment remains modest. And that is precisely what makes it so easy to disguise financial restructuring as AI transformation.

Nearly 245,000 tech jobs were cut globally in 2025. AI was cited as the cause in approximately 55,000 U.S. job cuts, according to Challenger, Gray & Christmas. But the actual causal chain — did AI replace these specific roles, or did the company use "AI transformation" as convenient framing for cost-cutting, post-pandemic correction, or shareholder appeasement? — is almost never examined.

Forrester's Predictions 2026 report went further: it predicted that half of AI-attributed layoffs would be quietly rehired, but offshore or at significantly lower salaries. According to Forrester, 55% of employers already reported regretting laying off workers for AI capabilities that did not yet exist.

Klarna is the textbook case. The company replaced 700 customer-service agents with an AI chatbot, announced it publicly as a success, then rehired human agents after quality collapsed. The CEO admitted they had let cost dominate the evaluation too heavily.

The most consequential workforce change in the technology sector is receiving comparatively little attention.

It is not mass layoffs.

It is a sustained contraction in hiring pipelines.

Indeed's data reveals the critical nuance: AI-mentioning postings are rising as a share precisely because overall postings are declining. The hiring market is not expanding. It is concentrating.

Federal Reserve Chair Jerome Powell characterised the current labour market as a "low-hire, low-fire" equilibrium. Companies are not aggressively cutting. They are simply not replacing.

The data on early-career roles is where this becomes visible.

That last finding is particularly revealing. Companies are not just posting fewer roles — they are stripping automatable duties from the roles they do post. The job itself is being redesigned around AI before a human ever applies.

The most rigorous evidence comes from the Stanford Digital Economy Lab. Erik Brynjolfsson and co-authors used high-frequency payroll data from ADP and found that early-career workers (ages 22–25) in AI-exposed occupations experienced 16% relative employment declines, while employment for experienced workers remained stable. The adjustments occur primarily via employment rather than compensation — meaning companies are not paying juniors less. They are simply not hiring them.

BLS data shows that employment for programmers — those who write code from specifications — fell 27.5% between 2023 and 2025. For software developers — those who decide what to build and how — it fell 0.3%. The more executable the work, the faster it disappears.

This dynamic is producing a structural tension.

LinkedIn's January 2026 Labour Market Report captured the collision in a single chart: record numbers of Computer Science graduates entering the market at the exact moment entry-level software engineering hiring hit record lows. The pipeline is producing more candidates than ever for positions that are evaporating.

Forrester found that Gen Z workers have the highest AI readiness score (AIQ) at 22%, compared to just 6% for Baby Boomers. The generation most capable of working productively with AI is the generation being locked out of the workforce by the elimination of entry-level positions.

The NACE Job Outlook 2026 survey found that approximately 45% of employers now characterise the job market for new graduates as merely "fair" — the most pessimistic rating since 2020. Employers project a marginal 1.6% increase in hiring for the Class of 2026 compared to 2025. Adjusted for the increasing number of graduates, this is a functional contraction.

The tasks that juniors used to learn on (code generation, financial modelling, baseline research, pitchbook preparation) are the exact tasks AI handles best.

This is not a temporary cycle.

This is a structural redesign of the entry point into the professional workforce.

The noise becomes signal only when you separate three phenomena that the media constantly conflates.

One counterpoint worth noting: Eurostat's EU-level data shows ICT specialist employment surpassed 10 million in 2024, reaching 5.0% of total employment, and the share is still rising. The U.S. tech-adjacent sectors have been contracting since 2022. The EU's specialist workforce has not. This does not mean Europe is immune — the entry-level squeeze is visible there too — but it means the picture is not uniformly bleak. It means there is still a window.

If you are reading headlines about AI layoffs and concluding that AI is making companies more productive with fewer people, you are reading the wrong story.

The companies announcing the largest layoffs are not the companies that have figured out AI. They are companies managing debt, correcting for over-hiring, or reallocating capital toward infrastructure bets that may or may not pay off. Oracle's 30,000 cuts have nothing to do with AI agents automating anyone's job. They have everything to do with a $108bn debt load and a cash crunch.

The companies that have actually figured out AI — the ones I wrote about last issue — look nothing like Oracle. They are 1-to-50-person teams built around AI from day one. They don't announce layoffs because they never hired the people in the first place.

On March 5, 2026, Bloomberg reported that Oracle was cutting thousands of jobs "for AI." The real story wasn't in the headline. It almost never is.

Fabio Lauria

CEO & Founder, ELECTE

This article is the second in a series on how AI is restructuring the operational architecture of businesses. For a deeper operational analysis, see the white paper: AI for European SMEs: The 2026 Playbook.

Sources:

Bloomberg, "Oracle Layoffs to Impact Thousands in AI Cash Crunch," March 5, 2026

Fortune, "Oracle under pressure from more than $100 billion in debt and massive layoffs," March 9, 2026

Wolf Richter, Wolf Street, "Despite 4 Years of Mass-Layoffs at Alphabet & Amazon, Headcount Rose in 2025, Nearly Flat with Peak, as Hiring Continued," February 9, 2026

CNBC, "Amazon layoffs: 16,000 jobs to be cut in latest anti-bureaucracy push," January 28, 2026

Amazon, Alphabet, Microsoft, Meta, Oracle — 10-K/Annual Report employee disclosures (2023–2025)

Federal Reserve Bank of New York, Liberty Street Economics, August 2025 regional business survey (AI adoption and labor effects)

Hampole et al., "Artificial Intelligence and the Labor Market," NBER Working Paper, February 2025 (revised September 2025)

Brynjolfsson et al., Stanford Digital Economy Lab / ADP payroll data analysis (early-career employment effects)

Challenger, Gray & Christmas, 2025 U.S. layoff data (55,000 AI-attributed job cuts)

Forrester Research, "Predictions 2026: The Future of Work" (rehiring prediction, AIQ data)

Indeed Hiring Lab, "January 2026 US Labor Market Update," January 22, 2026

LinkedIn, January 2026 Labour Market Report (CS graduate hiring data)

Revelio Labs, entry-level job posting decline data; AI-exposed task composition analysis

SignalFire, entry-level hiring decline study (2019–2024)

Handshake, early-career platform posting data

BLS / FRED, payroll employment series: Computer Systems Design (CES6054150001) and Information sector; JOLTS job openings

Eurostat, ICT specialists in employment (2024)

CNBC, "AI is not just ending entry-level jobs. It's the end of the career ladder as we know it," September 2025

IEEE Spectrum, "How AI Is Reshaping Entry-Level Tech Jobs," December 2025

NACE, Job Outlook 2026 survey

InformationWeek, "2026 tech company layoffs" tracker, March 2026