There is a certain kind of corporate honesty that makes people feel worse than silence. Oracle responded in June with its annual 10-K filing, which showed that the company had laid off about 21,000 workers over the past year. The company almost casually blamed AI for the cuts. Not even a hint. Given it a name. In a government document that can’t be taken back or changed by a PR firm.
As of May 31, 2026, the number of employees dropped from 162,000 to 141,000. Oracle spent almost five times as much on severance costs this year ($1.84 billion), which is a lot of money. Endings cost a lot of money.
The filing simply stated: “The adoption and deployment of AI technologies across our operations have resulted, and may continue to result, in reductions to our workforce.” That sentence should be read twice. Not job cuts because demand is falling, not a market correction, and not a reorganization of mergers. AI. It was written up in a formal SEC filing, so it’s not just something to talk about.
Oracle has changed the way its product development teams work so that they are “smaller, more agile, and productive.”Chairman and Chief Technology Officer Larry Ellison said that AI code generation means that fewer engineers are needed to do the work that used to need a lot of them. Oracle told its own shareholders that was the reason why tens of thousands of people no longer work there, so it’s not just a vague goal.
At the same time, Oracle promised to spend more than $70 billion to grow its AI and cloud infrastructure, mostly to help OpenAI and other customers. The business wants to get about $40 billion by selling debt and shares. During an earnings call in June, Ellison said that Oracle would build more cloud data centers “than all of its competitors put together.” It’s simple math: fewer people, more machines, and billions of dollars going the other way.
Some people don’t believe the story at face value. In May, Jensen Huang, CEO of NVIDIA, said that blaming AI for layoffs was “just too lazy.” He didn’t understand how a technology that has only been useful for a few months could already be putting so many people out of work. This year, Sam Altman said something similar about what he called “AI washing”—companies using AI as an easy way to explain cuts that might have happened anyway.
That doubt is based on something. Oracle has a little less staff than it did before it bought health records company Cerner for $28 billion in 2022. This makes me wonder if this restructuring is being driven by AI or if it is just a delayed fix for an overstretched expansion that is being said in the language of the moment.
Reports from India say that more than 12,000 jobs were cut. This made the situation more dangerous there. Oracle is also said to have pulled job offers made to recent graduates from IITs and NITs. These people had already made plans for work that didn’t happen. No matter what role AI played in those decisions, the results were always human.
The number of orders that Oracle has to fill has grown to $638 billion, which is more than four times what it was a year ago. Customer have “moved past the experiment stage with AI,” said Mike Sicilia, co-CEO.Demand is rising quickly. Headcount is going down. That doesn’t seem to be a contradiction to Oracle.
It is still not clear if other companies will follow Oracle’s lead and name AI directly in their regulatory filings, or if most will continue to play it safer. This year, Oracle made it clear that companies are no longer willing to just say things in their heads; they have to put them in writing. For the 21,000 people on the other side of that line, the outcome is likely more important than the words used.
