Last month Oracle let its investors and the SEC know that its “restructuring” costs would be $2.1 billion in 2026 – and swung an axe through its ranks. This week, 10,000+ staff* were let go via a 235-word email; three weeks after its CEO told investors it was going through “hyper-growth.”

*Oracle is not commenting on the precise number. Comments from insiders suggest a sweeping range of between 10,000 - 40,000.

The calls for help soon followed: “My entire team was laid off…” “Hi all, I just got laid off from Oracle due to business restructuring” “Anyone looking for a project manager or scrum master??? Just got laid off from Oracle.” 

The anecdotes came fast behind: The decade-long veteran who got a week’s pay for each year he worked there – less than three months’ pay from a firm reporting record profits. (“I hate this place. I really want to start my own sh*t,” he told friends); all the personal items left uncollected in offices…

“Access to your computer, email, voicemail, and files will be deactivated soon, and you will be unable to log into your computer. As a reminder, you are prohibited from downloading, copying or retaining (including emailing yourself) any Oracle confidential information.” - Oracle

(The mechanism is typically that the email arrives, passes are instantly locked, and if you were foolish enough after a decade of trustworthy value-building for a company to have left a favoured pot plant, family photo, or notepad on your desk, well, you better hope a friend’s pass is still active.) 

Tech layoffs: From AI to ZIRP

Corporate America and its technology sector in particular are no stranger to sweeping layoffs over the past few years. (The 16,000 job cuts at AWS in January have left some offices ghostly quiet: “Someone had to pay for the Melania documentary” is the morbid joke that apparently runs through them…)

Many large software companies hired aggressively during the zero interest rate (ZIRP) period. During the pandemic, demand for engineers also soared, as even companies that weren’t traditionally technology-focused scrambled to build out digital products and hire to support them. Both eras are over.

The layoffs come during a period in which the likes of AWS and Oracle are also spending colossal sums on AI; Oracle is ladling on debt to do so, also: In February it announced plans to issue $50 billion in debt – and within days raised $30 billion through a combination of investment-grade bonds and mandatory convertible preferred stock in an oversubscribed issuance. 

Oracle has not commented on the precise number of staff impacted. 

Many of the tech firms making similar sweeping layoffs are touting the need for renewed agility. Amazon, for example, said: “We’re convinced that we need…fewer layers and more ownership, to move as quickly as possible.”

The extent to which such job cuts are driven either by internal automation improvements (“we are leaning very heavily into AI ourselves. We have a thousand AI agents already live… yes we think AI is disruptive” Oracle’s co-CEO Mike Cicilia told analysts last month) or the need to improve liquidity given massive spending on AI hardware and energy is contested. 

Stepping back, however, as a team of researchers from Germany’s Mannheim University noted in October: “Employers systematically underestimate automation potential for their profession.

“Initial beliefs about the share of tasks that could be automated within the next decade are significantly lower than expert assessments…” 

The wave of white collar layoffs is likely to be far from over. 

nb: The Oracle layoffs come as academics emphasise that “flat and lean” organizations “cultivate more versatile AI generalists who transfer richer knowledge across firms” – that’s according to researchers at Chinese business school CKGSB and UT Dallas, writing in November 2025.  

“AI [productivity] spillovers differ fundamentally from traditional IT spillovers: while IT spillovers primarily arise from scale and process standardization, AI spillovers critically depend on the experimental and integrative environments in which AI knowledge is produced,” they said.

The link has been copied!