On March 31, 2026, Oracle executed the largest workforce reduction in its 49-year history: approximately 30,000 employees — 18% of the 190,000 global total — were laid off via emails sent at 6:00 AM. The stated goal: to free up between US$ 8 billion and US$ 10 billion in annual cash flow to fund the enterprise sector's largest AI infrastructure bet.

The Paradox: Record Growth and Historic Cuts

What makes Oracle's cuts exceptionally revealing is the financial context in which they occurred. In the third fiscal quarter of 2026, the company reported GAAP net income of US$ 3.7 billion — a 27% year-over-year growth. OCI (Oracle Cloud Infrastructure) grew 84%. Remaining performance obligations (RPO) expanded 325% to US$ 553 billion. Oracle is not cutting due to difficulties — it is cutting to fund a bet.

CFO Safra Catz described the action in a private call with analysts on April 2 as "a generational reallocation of capital from people-intensive consulting and legacy support to GPU-intensive AI infrastructure." Larry Ellison, in a separate meeting with Stargate investors, issued no public statement about the cuts — a characteristic choice.

What Is Stargate

Stargate is a joint venture among OpenAI, SoftBank, and Oracle announced with a US$ 500 billion commitment to AI infrastructure in the US. Oracle and OpenAI have an agreement to develop 4.5 gigawatts of additional Stargate data center capacity — a partnership that exceeds US$ 300 billion between the two companies over five years. To fund its share, Oracle raised US$ 30 billion in debt and convertible preferred stock in February 2026.

Who Was Affected

The cuts focused on three areas: Cerner/Oracle Health (acquired for US$ 28 billion in 2022, considered heavy on human implementation), Oracle Cloud Infrastructure in non-AI roles, and middle management replaced by automated workflow tools. Stargate data center and AI infrastructure teams were explicitly spared — Oracle was simultaneously hiring in these areas while cutting elsewhere.

The Defining Pattern of 2026

Oracle's cuts are the single largest data point in a documented pattern in 2026: Accenture (11,000), Amazon (16,000 corporate), Microsoft (15,000 in 2025), Meta (8,000 in February 2026). In all cases, workforce reductions are announced concurrently with massive AI investment commitments. The pattern has a simple name: payroll-to-capex — transferring capital from human costs to computational infrastructure costs.