Oracle Is Laying Off Thousands While Filing Thousands of H-1B Petitions. That's the Point.

One hand signs the pink slips. The other files the visa petitions. Same company, same quarter. Oracle’s workforce strategy over the past two years is perhaps the starkest illustration of how Big Tech really thinks about labor — not as people, but as line items to be optimized.

The Two-Door Company

Oracle has conducted multiple rounds of mass layoffs from 2024 through 2026, primarily gutting legacy on-premises teams as the company pivots hard toward cloud infrastructure. Standard restructuring story. What makes it uncomfortable is that during the same period, Department of Labor records show Oracle consistently filing thousands of H-1B visa petitions each year.

It looks like a contradiction. It isn’t. The exit door and the entrance door lead to different hallways. The layoffs target senior engineers and middle managers tied to older product lines. The H-1B hires flow into cloud, AI, and data engineering roles. Same company, completely different talent pipelines. Corporate restructuring and workforce arbitrage, running in parallel.

Why H-1B Workers Are Structurally Attractive

The reasons Big Tech leans on H-1B hiring aren’t subtle. They’re structural.

Cost control. H-1B salaries aren’t necessarily lower on paper, but the leverage equation is different. Visa sponsorship acts as a de facto retention mechanism — switching employers means restarting an immigration process that can take years. That kills bargaining power. Companies get a workforce that’s less likely to negotiate aggressively or walk.

Predictability. When your employee’s legal right to remain in the country depends on continued employment with you, the relationship goes well beyond a standard offer letter. Pushing back on working conditions or demanding raises carries existential risk for the worker. That asymmetry benefits the employer in ways that never show up in compensation data.

Talent access. This one is real, and it’s worth acknowledging honestly. For certain specializations — large-scale distributed systems, cloud-native architecture, advanced ML infrastructure — domestic supply genuinely doesn’t meet demand. The global talent pool matters.

The problem is that all three reasons operate simultaneously, and companies aren’t exactly transparent about which one is driving any given hire.

This Isn’t an Oracle Problem

Every major tech company runs this same playbook. Google, Meta, Amazon, Microsoft — pick one. Between 2023 and 2025, Big Tech collectively laid off over 300,000 workers. During the same window, H-1B application volumes actually increased.

The resentment in US tech worker communities has been building for years, and it’s not hard to see why. When a senior engineer with a decade of experience gets cut and the role reopens as a junior H-1B position, the “global talent access” argument starts to ring hollow. On Hacker News and Reddit, the discourse has shifted from policy debate to something closer to raw anger. The charge: H-1B has been co-opted from a talent program into a cost-reduction tool.

Policy Caught in the Middle

The Trump administration has pushed toward stricter H-1B enforcement — higher prevailing wage requirements, tighter specialty occupation definitions, more scrutiny on petitions. But Big Tech’s lobbying apparatus is formidable, and the national competitiveness argument (“we need STEM talent to beat China”) remains politically potent on both sides of the aisle.

The result is a system that satisfies nobody. Companies still find the process slow and uncertain. Domestic workers see it as a threat to their livelihoods. And H-1B holders themselves are trapped in a structure where their immigration status is functionally controlled by their employer — a dynamic that raises its own serious fairness questions, regardless of where you stand on the broader debate.

The Real Discomfort

What makes Oracle’s case sting isn’t the simultaneity of layoffs and hiring. It’s what the pattern reveals about how the industry values people. The message encoded in the numbers is blunt: workers are interchangeable components in a cost optimization function. Swap out the expensive ones, plug in the cheaper ones, call it transformation.

If you work in tech, this is worth sitting with for a moment. What determines your market value — your skills, or your visa status and country of residence? If that question makes you uncomfortable, you’re probably reading the situation correctly.