Oracle disclosed in its annual financial filing that it cut 21,000 employees in the past 12 months. That is 13% of its entire global workforce. The company spent $1.84 billion on severance in fiscal 2026, compared to $374 million the year before. Oracle is posting record profits and has $553 billion in contracted future revenue, so this is not a company under financial pressure. It is choosing to free up cash to build AI data centers.
Oracle is not alone. This is now a documented, named trend across the tech sector. More than 120,000 tech jobs were cut in the first five months of 2026, a 65% jump over the same stretch in 2025. AI was the most-cited reason for layoffs across all industries for three consecutive months, according to Challenger, Gray & Christmas. In May alone, AI was cited for 40% of all job cuts announced that month.
The companies doing the cutting are not the ones you would expect to be in trouble. Amazon, Meta, Google, Oracle, Salesforce, PayPal, Intuit, Cisco: all of them are posting strong financial results. Google's cloud business just crossed $20 billion in quarterly revenue for the first time. Meta cut 8,000 people while telling investors it was necessary to fund AI investment. Salesforce said its AI system now handles enough support cases that the company simply no longer needs to fill those roles.
There is a clear trade being made. Google, Amazon, Microsoft, and Meta together plan to spend $725 billion on AI infrastructure in 2026, up 77% from last year. That money has to come from somewhere, and payroll is one of the largest controllable costs any company has. Cutting human roles and redirecting that budget toward AI systems is, from a financial standpoint, a straightforward calculation.
The roles being cut follow a consistent pattern across every company on this list. Customer support, middle management, internal operations, finance, legal support, data entry, quality assurance. Cloudflare's CEO described the people laid off as mostly "measurers," meaning the managers, auditors, and coordinators who track and report on the work others do. Google cut more than a third of its managers overseeing small teams. Coinbase flattened its entire structure to five layers below the CEO. This is not about eliminating one department. It is about removing the human layer that exists between decisions and execution.
Here is the part that should concern readers outside the tech industry: this is no longer contained to software companies. AI-attributed layoffs have spread into finance, logistics, consulting, retail, and manufacturing. Law firm Baker McKenzie cut up to 1,000 employees in roles like research, secretarial, and know-how functions. McKinsey cut back-office and support roles. Citigroup is working toward a reduction of around 20,000 people. The roles being eliminated at these firms look exactly like the roles being eliminated at Oracle or Amazon.
There is a reasonable counterargument, and it deserves to be stated clearly. Some of these layoffs are almost certainly "AI washing": companies using AI as a publicly acceptable reason for cuts that are really about undoing pandemic-era overhiring or improving profit margins before an investor call. Challenger's data shows that market conditions and restructuring still lead AI as stated reasons year-to-date. Unemployment claims have not spiked in line with the announcements.
But the honest answer is: both things are true at once. Some of these cuts would have happened anyway, and some of them are genuinely caused by AI tools doing work that people used to do. Salesforce's Agentforce product is handling support cases that previously required human agents. Amazon's CEO said publicly in 2025 that AI would reduce the total corporate workforce, and then Amazon cut 30,000 corporate jobs in less than six months.
What does this mean if you are not in tech? The roles most exposed are the ones that involve processing, coordinating, or reporting on information without adding judgment or client relationships. If a role's primary output is a document, a report, a ticket, or a status update, AI can already do a version of that work. The roles that are growing at the same companies doing layoffs are the ones requiring judgment, accountability, and direct relationships with clients or regulators. That is where the line currently sits.