On the morning of March 31, employees at Oracle across the United States, India, Canada, Mexico, and Uruguay opened their email to find a termination notice from "Oracle Leadership." It arrived as early as 6am. There was no prior conversation with a manager, no warning from HR, and no severance discussion in person. The email informed recipients that their roles had been eliminated "as part of a broader organisational change" and that the same day would be their last. Investment bank TD Cowen estimates between 20,000 and 30,000 employees were cut: roughly 18% of Oracle's global workforce of 162,000 people. India alone lost approximately 12,000 positions.
The reason given, when any reason was given at all, pointed to AI. Oracle has committed to an aggressive infrastructure buildout estimated at $156 billion in capital spending, and analysts at TD Cowen calculate that cutting this many workers could free between $8 billion and $10 billion in incremental cash flow. The logic is straightforward: the company needs money for AI data centres, and human labour costs money. The workers went to fund the machines.
What makes this particular case harder to dismiss than most corporate restructuring stories is the financial context. Oracle did not cut these jobs from a position of distress. The quarter before the layoffs, the company posted a 95% jump in net income, reaching $6.13 billion. Its remaining performance obligations, a forward-looking indicator of contracted revenue, stood at $523 billion, up 433% year over year. This is not a company that could not afford its workforce. It is a company that chose to redirect the money those workers generated toward something it values more.
Oracle is not alone in this calculation. Analysts tracking the broader tech sector have noted that large technology companies are now cutting close to 1,000 jobs per day collectively, with AI infrastructure spending cited as the primary driver. The pattern is consistent: record profits, record AI investment commitments, and workforce reductions that free up the capital to bridge the gap. The workers are not being let go because they failed to perform. They are being let go because their salaries are now competing directly with GPU capacity on the balance sheet.
The 6am email detail has attracted its own attention, and for reasons beyond optics. It reflects something about the decision's character: the same efficiency logic that is being used to justify replacing workers with automated systems was applied to the act of telling those workers they were gone. A standardised message, sent at scale, at the moment it became administratively convenient. Whether that constitutes a failure of decency or simply the logical endpoint of treating labour as a cost variable is a question different observers are answering very differently.
What is not in dispute is the direction. Oracle has placed its bet on AI infrastructure and has priced that bet in jobs. The $8 to $10 billion in freed cash flow will go toward data centres that will serve AI workloads for customers who are, in many cases, running the same calculation Oracle just ran: what human roles can be made redundant, and how quickly.