Oracle lays off 30,000 employees to fund its AI infrastructure bet

Oracle is laying off somewhere between 20,000 and 30,000 employees worldwide, making it the largest single tech layoff in the past year. The cuts hit on March 31, with many employees in India reportedly receiving termination notices in a 6 AM email. Oracle employs roughly 162,000 people globally, so this would represent up to 18% of its workforce.

The stated reason is straightforward: Oracle is redirecting spending toward AI infrastructure. The company has been pouring capital into data center buildout to compete with Amazon, Microsoft, and Google in the AI cloud race. TD Cowen analysts estimated back in January that cutting 20,000 to 30,000 employees could free up $8 billion to $10 billion in incremental free cash flow. Oracle is effectively trading headcount for GPU clusters.

The departments hit span engineering, cloud operations, sales, and cybersecurity teams. Reports on the India impact vary: some outlets say 2,500 roles affected, others cite figures as high as 12,000. The discrepancy likely comes from Oracle not issuing an official number, leaving media to piece together estimates from affected employees and internal sources. A second round of cuts is reportedly being considered.

Oracle’s stock moved in premarket trading on April 1, though the direction tells you everything about the current AI narrative. Investors are rewarding the cost-cutting signal while the broader market wrestles with whether AI infrastructure spending will ever generate returns proportional to the capital being poured in. The company’s capital expenditures have surged as it builds data centers to handle AI workloads, and Wall Street has been pressuring management to show a clearer path to profitability on those investments.

This is not an isolated event. Amazon, Microsoft, Intel, and others have all conducted similar workforce reductions tied to AI pivots over the past two years. The pattern is consistent: legacy software and operations roles get cut, AI and infrastructure roles get funded. The difference with Oracle is scale. Thirty thousand jobs in one round is a number that even by 2026 standards stands out.

For employees, the message is blunt. The severance packages being offered require employees to accept terms through a digital signature portal, a detail that has drawn criticism for removing human intermediation from an already painful process. Some reports describe workers who had been with the company for over a decade finding out they were done through an automated email.

Oracle’s leadership has maintained that the AI investment will pay off over time. The company holds contracts to provide cloud infrastructure for major AI workloads and has positioned itself as a key partner for enterprises running large-scale training and inference. Whether that bet generates enough revenue to justify the human cost is a question the company will need to answer on its next earnings call.