SUSE told roughly 1,900 employees on Monday that about 190 of them are surplus, and credited automation and AI for the arithmetic. The calendar says something less flattering: the company tried to buy those exits voluntarily in September, could not find enough takers, and will not show the survivors its strategy until November 16 — two weeks after the names come down.
The memo went out Monday from chief executive Dirk-Peter van Leeuwen, the former Red Hat executive who took the job in March 2023, according to The Register, which obtained it. The message announced a reduction of about a tenth of the global workforce and did not say which roles, teams, functions or countries are affected. People caught in the cut are to be informed by Wednesday. Employee representatives are to be nominated Thursday. One-to-one meetings begin early next week. The fiscal 2027 strategy gets presented on November 16.
That is the sequence. Two days of notices, one day to elect the people who will sit opposite management, then the meetings, then the plan. A company that will not say which jobs are going is asking the people who stay to trust a strategy it has not shown them yet.
Dirk-Peter van Leeuwen — The Memo Blames AI for a Number Somebody Chose Months Ago
The reasons in the memo are broad enough to cover almost any decision. Automation and AI are said to be fundamentally changing how software companies operate. Modern tooling is already changing daily workflows, teams are said to move faster with less operational overhead, and reporting lines are getting flatter. The need for certain existing roles has fallen sharply while other roles will evolve over the next few years, and spending is being pushed toward priority areas. The organizational design and cost base SUSE runs on today, the memo argues, will not carry it forward.
Now put that next to September. heise online reports that SUSE asked its long-serving staff in September whether they would accept voluntary separation, and that too few volunteered — which is why the cuts are now involuntary. Nine weeks separate the ask from the memo. Whatever automation is doing inside SUSE, the headcount target existed while the company was still hoping people would leave on their own.
Voluntary exits that fail are not a discovery about technology. They are a budget target that did not fill itself.
The 190 — No Names, No Roles, No Countries, Just a Wednesday Deadline
What SUSE has said about how it chose is that managers looked at where people could develop, which tasks they could stop doing, and which they should take on more of. What it has not said is anything a person could check. No list of functions. No breakdown by office. No count from engineering versus support versus sales. The company’s statement that the cuts reflect organizational need rather than individual talent is generous in tone and empty in content, because every redundancy letter in history says the same thing.
Consider the shape of the week from inside one of those 190 lives. Monday you learn the number. Wednesday you learn whether you are the number. Thursday your colleagues nominate the representatives who will negotiate over your exit. Some time next week you sit in a room and hear the details. On November 16 the company explains the future it is building without you. Whatever severance is on offer, the order of events tells you which party had a plan and which party is being managed.
This is a 34-year-old company with long-tenured staff, and the September request was aimed specifically at the people who had been there longest. Those are the engineers who remember which customer runs which cluster, the support hands who have been paged at three in the morning for a decade, the people whose names sit in the commit history of the products SUSE sells. Loyalty is the asset the open source pitch runs on. It is also the cheapest thing to spend when a spreadsheet needs a smaller denominator.
You do not build a 34-year reputation on loyalty and then cash it in on a nine-week calendar.
EQT — The Exit Math Does Not Need a Chatbot to Get to 190
SUSE is owned by EQT, the Swedish buyout group that took it private in 2023, when the business was valued at just under $3 billion. Earlier this year there were reports that EQT was exploring a sale, with a price of up to $6 billion under discussion, as heise notes. The 1,900 headcount figure comes from EQT’s own portfolio page. A ten percent reduction at an enterprise software company with a fund owner and a live exit question is not a mysterious event. It is a cost line moving.
The wider week makes the pattern easy to see. ZDNET notes Microsoft offering buyouts to 60,000 workers and Squarespace cutting 19 percent of its staff in a message that never names AI at all. ZDNET’s reporters also make the point the industry would rather skip: from the outside, a layoff that is strategically blamed on AI and a layoff that was planned before AI are nearly impossible to tell apart. The label is doing work the evidence has not done.
Private equity does not need a technology story to want a smaller company. It needs one the market will repeat.
The AI-Mekna — Nothing in the Memo Counts What the Software Actually Replaced
There is no system named in the memo. No deployment described. No function mapped to a tool. No before-and-after metric, no ticket volume, no headcount-per-team comparison, nothing that would let an employee or a customer test the claim that automation made 190 roles unnecessary. That absence is the story. Companies genuinely reorganizing around AI can usually point at the thing doing the work — a specific product, a specific workflow, a specific team that shrank because a specific job stopped existing.
AI-mekna is real, and it is also the most convenient noun in corporate communications right now. Both of those things can be true in the same memo, which is exactly why this one deserves a second read rather than a headline. Automation does not sit in a boardroom and pick a number. A chief executive with a cost target does, and the word arrives afterward to make the number sound like weather.
The Ledger — 190 Now, and the Rehire Bill Comes Due Around 2029
The bill for this kind of cut rarely arrives on schedule, which is why it gets missed. ZDNET cites research projecting that companies will end up rehiring roughly 30 percent of the workers they tried to replace with AI — at higher cost — by 2029, unless they overhaul how they plan their workforce. Rehiring costs more than retaining because the institutional knowledge walked out with the badge, and the contractor who replaces it has to be taught the customer’s environment from scratch.
For a company selling enterprise Linux support, that knowledge is the product. Reliability is not a benchmark score; it is the person who recognizes the pattern in a stack trace at 2 a.m. because they saw it in 2019. Cut a tenth of a 1,900-person company and you have not removed a tenth of its overhead. You have removed a tenth of its memory, spread unevenly, concentrated exactly where tenure was longest.
Watch three things and the truth will surface on its own. The November 16 strategy will show whether this is a real reallocation or a costume. The company’s next disclosure will show whether the shrinking functions match the story it told. And the 2027 numbers will show whether AI shows up at SUSE as revenue it sells or only as cost it saves. One of those is a business. The other is a press release with a balance sheet attached.
The simplest reading is the one the memo avoids. A fund-owned company, nine weeks after a voluntary exit that failed, found a word that made 190 redundancies sound inevitable. Inevitable things do not need severance negotiations, employee representatives, or one-to-one meetings scheduled for next week. People do.
Sources: The Register, heise online, ZDNET, EQT.