Oracle’s massive layoffs: 30,000 jobs on the chopping block

What happens when a tech giant bets big on AI but can’t find anyone willing to foot the bill? Oracle’s about to find out the hard way.

The database giant is planning to slash up to 30,000 jobs as US banks are backing away from financing the company’s massive AI data center expansion. That’s not a typo. We’re talking about potentially cutting one in four employees to free up cash for Oracle’s AI ambitions.

The Numbers Don’t Lie

Here’s where it gets ugly. Oracle needs $156 billion in capital expenditures to build out its AI data center infrastructure.

For context, that’s more than the GDP of Ukraine. The company had committed to building data centers for OpenAI, but now multiple US banks have pulled out of lending deals over the past few weeks.

The proposed layoffs would affect between 20,000 and 30,000 employees. That would generate $8 billion to $10 billion in free cash flow for Oracle. It’s a brutal trade-off, but the math is pretty straightforward when you’re staring down a $156 billion funding gap.

Both equity and debt investors have raised questions regarding Oracle’s ability to finance this buildout.

This wouldn’t be Oracle’s first rodeo with mass layoffs either.

The company cut around 10,000 jobs in late 2025 as part of a $1.6 billion restructuring plan. But 30,000? That would be the largest in the company’s recent history by a mile. To put this in perspective, Oracle’s current workforce sits at approximately 160,000 employees globally. Cutting 30,000 positions means eliminating nearly 19% of its entire workforce. That’s not trimming fat – that’s amputating limbs to keep the body alive.

When the Money Dries Up

The financing crunch is already hitting Oracle’s customer relationships. Multiple data centre leases that were under negotiation with private operators couldn’t secure financing, which meant Oracle couldn’t lock down the capacity they needed through leasing arrangements.

Multiple Oracle data-centre leases that were under negotiation with private operators struggled to secure financing, in turn preventing Oracle from securing the data-centre capacity via a lease.

It’s a domino effect.

Banks won’t lend, which means Oracle can’t lease, which means they’re scrambling for alternatives. And those alternatives are getting creative in ways that probably make their enterprise customers uncomfortable. Oracle is now requiring clients to help build infrastructure. Translation? They’re asking customers to shoulder part of the cost.

The company is also exploring something called “bring your own chip” (BYOC), where new customers have to supply their own hardware. It’s basically Oracle saying “we’ll run your workloads, but you need to bring your own silicon.”

The bank pullout isn’t happening in a vacuum. Financial institutions are reassessing the risk-reward ratio of massive AI infrastructure investments after seeing mixed results from early AI deployments across various industries.

The initial AI hype of 2023 and early 2024 has given way to more cautious evaluation of actual returns on investment.

What This Means for Your Wallet

This whole mess highlights something nobody wants to talk about in the AI space.

Building the infrastructure to support large language models and AI workloads at scale is incredibly expensive. We’re not just talking about a few servers in a closet. These AI data centres need specialised cooling systems that can cost $50 million per facility, power requirements that exceed 100 megawatts per centre, and hardware that costs more per rack than most people’s houses. Oracle’s commitment to OpenAI alone represents a staggering financial obligation. The $156 billion figure isn’t just a rough estimate – it’s what Oracle needs to fulfill its AI infrastructure promises. That’s roughly equivalent to building 300 large-scale data centres simultaneously.

And honestly? Oracle isn’t alone in this crunch either.

Amazon just laid off 16,000 workers as part of its own AI restructuring plan. The AI boom is real, but so are the infrastructure costs. Google has spent over $31 billion on AI infrastructure in 2024 alone, while Microsoft allocated $44 billion for similar investments. The difference is those companies had the cash reserves and credit lines to make it happen. Thing is, Oracle doesn’t.

The company is also considering selling its healthcare software unit, Cerner, which it acquired for $28.3 billion in 2022. That’s another sign of how desperate they’re for cash. You don’t sell a unit you paid nearly $30 billion for two years ago unless you’re really feeling the squeeze. The Cerner acquisition was supposed to be Oracle’s entry into the massive healthcare IT market, but now it’s looking like expensive dead weight.

Bad News for Canadian Tech Workers

While Oracle hasn’t broken down where these layoffs will hit geographically, the company has significant operations in Canada.

Including development centres in Vancouver, Toronto, and Ottawa. Oracle’s Canadian workforce includes approximately 3,500 employees across software development, sales, consulting, and support functions.

Canadian tech workers at Oracle are likely watching this closely, especially those working on cloud infrastructure and AI-related projects. The Vancouver office, which focuses on cloud applications development, and the Toronto centre, which handles enterprise software solutions, could be particularly vulnerable given their connection to the areas Oracle needs to streamline.

Look, the ripple effects could extend beyond Oracle too. If one of the big players in enterprise software is struggling to secure financing for AI infrastructure, it raises questions about how sustainable the current AI expansion really is. Canadian tech companies that have been riding the AI wave might need to reassess their own expansion plans and funding strategies.

For Canadian companies that rely on Oracle’s cloud services, the “bring your own chip” model and client-funded infrastructure could mean higher costs and more complexity. A mid-sized Canadian manufacturing company that currently pays $2 million annually for Oracle cloud services might face additional hardware costs of $500,000 to $1 million under the new BYOC model.

The Canadian tech sector has been relatively insulated from major layoffs compared to Silicon Valley, but Oracle’s cuts could change that dynamic. With tech unemployment in Canada currently at 2.1%, significantly lower than the 4.5% rate in major US tech hubs, any substantial Oracle layoffs would represent a meaningful shift in the local market.

Brutal.

Following the Money Trail

Oracle’s current situation stems from a fundamental mismatch between AI ambitions and financial reality.

The company’s annual revenue for 2024 was $53 billion, meaning the $156 billion AI infrastructure investment represents nearly three years of total revenue. That’s not sustainable math, especially when you consider Oracle still needs to fund its existing operations, research and development, and shareholder returns. The company’s debt-to-equity ratio has already climbed to 1.8, well above the industry average of 0.9 for enterprise software companies.

Adding another $156 billion in debt obligations would push that ratio to potentially unsustainable levels, which explains why banks are backing away from the deals.

Oracle’s stock price has reflected this uncertainty, dropping 12% since news of the financing troubles began circulating in investment circles.

The company’s market capitalisation has fallen from $485 billion in early 2024 to approximately $427 billion currently, representing a loss of nearly $60 billion in shareholder value. The timing couldn’t be worse for Oracle’s expansion plans. Interest rates remain elevated compared to the near-zero rates that fueled tech expansion in previous years. Corporate lending rates for large infrastructure projects are currently averaging 8.5% to 11%, making the carrying cost of a $156 billion investment potentially catastrophic for long-term profitability.

Playing Catch-Up in a Brutal Race

Oracle’s desperate moves come as competitors are pulling ahead in the AI infrastructure race.

Amazon Web Services controls approximately 32% of the global cloud infrastructure market, while Microsoft Azure holds 23%. Oracle’s cloud infrastructure market share sits at just 2%, making its massive investment seem like a Hail Mary attempt to remain relevant. The pressure intensified when OpenAI began diversifying its infrastructure partnerships. While Oracle secured some high-profile contracts with the AI company, OpenAI has also signed deals with Microsoft and Google, reducing Oracle’s strategic importance. This diversification strategy by AI companies has left infrastructure providers like Oracle competing for smaller slices of a very expensive pie.

But here’s the thing nobody’s talking about enough.

The broader enterprise software market is also shifting toward more cost-conscious purchasing decisions. After two years of aggressive AI spending, corporate IT departments are demanding clearer ROI metrics before committing to new AI infrastructure investments. This trend has made Oracle’s client-funded infrastructure model even more challenging to sell. Companies aren’t just throwing money at AI projects anymore (which, honestly, they probably should’ve stopped doing earlier).

What Happens Next?

Oracle expects to raise $45 billion to $50 billion in 2026 to build additional cloud infrastructure capacity.

Even with the potential $10 billion in savings from layoffs, they’re still looking at a massive funding gap of over $95 billion. The company will likely need to pursue a combination of asset sales, strategic partnerships, and potentially even equity raises to bridge this gap. The company hasn’t released an official statement about the layoffs yet, which isn’t surprising. Mass layoff announcements are usually carefully orchestrated PR exercises, not leaked through investment bank research reports. However, Oracle’s silence speaks volumes about the sensitivity of the situation and the potential impact on employee morale and customer confidence.

Industry analysts expect Oracle to announce the layoffs within the next 60 days, likely timing the announcement to coincide with quarterly earnings to minimize separate negative news cycles.

The layoffs will probably be structured as voluntary buyouts initially, followed by involuntary cuts if not enough employees accept the packages. But the writing is on the wall. Oracle bet big on AI infrastructure, assumed the funding would follow, and now they’re scrambling when banks are getting cold feet about lending hundreds of billions for data centres that might not generate returns for years.

The tech industry has seen this pattern before during the dot-com crash of 2001 and the financial crisis of 2008. Massive expansion plans, easy money assumptions, then reality hits and the layoffs begin. Oracle’s situation is just playing out on a scale that makes previous tech layoffs look quaint.

The broader implications extend beyond Oracle to the entire AI infrastructure sector (for better or worse). If established players like Oracle can’t secure financing for AI expansion, it raises serious questions about the sustainability of current AI growth projections.

What This Means Going Forward

The industry might be approaching a reckoning where AI promises need to align more closely with financial realities.

Thirty thousand jobs to chase an AI dream that might not pay off. That’s either visionary leadership or a spectacular miscalculation.

We’ll find out which one in the next few quarters, but the early signs aren’t encouraging for anyone betting on Oracle’s AI future.