Analysis
What Changed
Pulse covered Oracle cutting 21,000 jobs in June to help fund its debt-fueled AI data-center buildout. Two months later, the pattern is repeating: Oracle is drawing up another round of layoffs this month, with some teams facing double-digit percentage cuts, according to Yahoo Finance, which reports the company has asked managers to identify affected employees ahead of the September 1 start of its second fiscal quarter. What's new since June isn't the strategy -- it's the scale of the financial commitment behind it, now fully quantified.
The Debt Picture
Oracle's total debt has reached $129.5 billion, alongside $260 billion in data-center lease commitments -- a scale of obligation few enterprise software companies have ever carried:
- FY26 capital expenditure -- $55.7 billion, up sharply from $21.2 billion the year before
- Cash outflow beyond internal generation -- $23.7 billion
- Debt raised in FY26 -- $43 billion, plus $5 billion more through stock sales
- Expected additional raise, year ahead -- roughly $40 billion more in borrowing and equity issuance
That's a company financing an AI bet with debt at a scale that leaves little room for a slowdown in the demand it's betting on.
The Other Half of the Story
The layoffs are landing alongside expansion, not contraction, in Oracle's actual business. Oracle and AWS announced that Oracle AI Database@AWS -- running on Exadata-class infrastructure -- is now generally available across 22 regions spanning Asia Pacific, Europe, and the Americas, according to PRNewswire's release, with a new long-term strategic collaboration agreement signed between the two companies. It's the same multicloud playbook Oracle ran with its expanded Google Gemini partnership in July -- letting enterprises keep their Oracle databases while using AWS or Google for broader cloud infrastructure, rather than forcing an all-or-nothing migration onto Oracle's own cloud.
The underlying business is genuinely growing: Oracle posted 17% revenue growth in fiscal 2026, with its cloud infrastructure segment up 77% on surging demand for AI-workload compute. Competitors AWS, Microsoft Azure, Google Cloud, and newer AI-cloud specialists like CoreWeave and Nebius are all racing for the same enterprise AI-infrastructure spend.
The Counterweight
A company can be growing revenue and still be over-levered if its debt-service costs outrun the returns on the infrastructure that debt is funding. $129.5 billion in debt plus $260 billion in lease commitments is a scale of obligation that assumes AI-compute demand keeps growing at its current pace for years -- if enterprise AI spending slows before Oracle's data centers are fully utilized, the layoffs happening now will look like the easy part of the adjustment, not the hard part.
Watch Oracle's next earnings report for whether cloud infrastructure growth holds above 70%, and whether the AWS and Google multicloud deals meaningfully offset the capital Oracle is borrowing to build its own data centers -- that ratio is what determines whether this debt load pays for itself.
