The Austin, Texas-based company disclosed in an annual regulatory filing earlier this year that its total workforce decreased by roughly 21,000 employees from 2025, costing the company roughly $2.1 billion mostly through severance packages and contract terminations.
Those cuts affected 741 Seattle-area employees between August 2025 and March, which, coupled with this week’s layoffs, means 1,100 Oracle workers in Seattle have been let go since last summer.
In a quarterly regulatory filing, Oracle addressed its 2026 restructuring plan and said it took further action after the fiscal year ended on Aug. 31, costing an additional $700 million and causing this week’s layoffs.
Oracle did not immediately respond to a request for comment.
Morally perhaps, but revenue wise Oracle is still increasing its profitability in its core business. Here’s the relevant excerpt from Oracle’s the June 10-K filing. Red circled is 2025 revenue and Blue circled is 2026. They grew total revenue $10 billion in a year.
If you’re thinking of Oracle’s (small and medium sized) database business and Java licensing as falling out favor, I think you’re right. However, thats not where Oracle makes much of its money these days. I also think you’re right when you say folks are migrating away for small and medium sized DBs. The cost for using Oracle DB on-prem is usually not worth it. Really good open source DB software like MySQL (MariaDB) or Postgres can handle that load with a tiny fraction of the operational costs. Additionally big enterprise customers are using cloud services from MS, AWS, and GCP for managed database solutions for those sized workloads.
However, on the big DB side Exadata is still in use and I’m not seeing companies migrate away from it. Oracle’s big revenue drivers these days is big enterprise EPM and ERP software as well as Oracle cloud hosting contracts. Additionally, OpenAI bought up nearly every spare CPU in Oracle Cloud to expand its GPT hosted offerings. Its risky though because OpenAI is paying Oracle with other people’s money, so if OpenAI’s money dries up, Oracle is going to be left with massively huge unpaid bills.
I think I may be the same age as you are, but Oracle buying Sun Microsystems and what Oracle did with Java is lots of ancient history. Yes a few years ago there was an additional squeeze on existing Oracle Java licensees, but like you said (and I agree with) lots of companies moved away from Java more than a decade ago. An entire generation of new IT professionals has been born, grew up, learned IT, and have no memory of the this debacle of the Sun purchase.
The huge risk to Oracle is the gamble they’re taking on OpenAI. Oracle is dumping MASSIVE amounts of borrowed money into a big datacenter build out and they’re doing it with a promise from OpenAI that they’ll buy up the hosting capacity. Oracle is getting its money buy issuing corporate bonds. Oracle’s bond rating has dropped to a horrible BBB- rating. One more drop and its considered "junk bonds".
If the AI bubble pops in the next year or so, Microsoft, Amazon, and Google will feel a sting, but move on. Oracle will collapse and fall over dead.
Kind of… MySQL is really not comparable to Postgres, at least as MySQL is generally used. It’s commonly set up to sacrifice ACID properties in favor of performance. When configured to go as full ACID as it can, its performance and robustness are inferior to Postgres. MySQL/Maria is great in its little niche, though. And I suppose the same could be said about Mongo.
From your lips to God’s ears.
I was forced to use Oracle’s 10g in college. That was my only interaction with Oracle and I’ll still hate them as long as I live. I’m rooting for OpenAI to blow up just to screw Oracle.