Cloudflare cut more than 1,100 jobs while its engineering ranks kept growing, and Matthew Prince is saying the quiet part plainly: AI is changing the shape of companies before it destroys them outright.
The numbers don't fit the usual layoff story. On May 7, Cloudflare said it would cut about 20% of its workforce, more than 1,100 people, on the same day it reported Q1 2026 revenue of $639.8 million, up 34% from a year earlier. Shares still fell sharply after the announcement. Investors didn't see a clean growth story. Employees certainly didn't see one. Prince's argument is blunter than that: the company isn't shrinking so much as changing what kinds of work it wants humans to do.
Business Insider's latest reporting added the part that makes this more than another AI layoff memo. Cloudflare employed 1,308 engineers in December, according to LinkedIn data analyzed by BNP Paribas and cited by Business Insider. That number has since climbed to 1,894, a 45% jump. Prince confirmed the trend to the outlet. So yes, the company cut deeply. It also kept hiring the people who build.
That is the real story here. Cloudflare didn't present the cuts as a punishment for weak performance or a scramble to survive. Prince and co-founder Michelle Zatlyn said the restructuring was tied to the company's move into what they called the agentic AI era. Very few engineers or customer-facing sales staff were affected, Prince wrote in response to criticism on X, and Cloudflare would keep hiring in those roles like it did when it was a startup.
Prince has been using an old Peter Drucker frame to explain the split. A company needs builders, sellers and measurers. Builders make the product. Sellers bring in customers. Measurers track, audit, coordinate, report and manage the machinery around the business. AI, in Prince's view, is falling hardest on that third group because much of that work is repetitive, rules-based and built around moving information from one place to another.
You don't have to like the conclusion to see why founders are paying attention. Cloudflare isn't a two-year-old AI startup with a thin payroll and a pitch deck. It's a public infrastructure company founded in 2009, with thousands of employees and nearly $640 million in quarterly revenue. When a company like that says its old staffing model no longer fits its internal productivity data, you should treat it as a signal, not a slogan.
The company's own figures explain why Prince felt able to move this fast. In the memo reported by Business Insider and the San Francisco Chronicle, Cloudflare said internal AI usage had risen more than 600% in three months. Employees across engineering, HR, finance and marketing were running thousands of AI agent sessions each day. That is not a chatbot experiment sitting in a side channel. It is operating software moving through the actual company.
The bill is not small. The San Francisco Chronicle reported that Cloudflare expects $140 million to $150 million in restructuring charges, mostly for severance, benefits and related costs, with most of the charge landing in the second quarter and the plan substantially complete by the end of the third quarter. Departing U.S. employees are set to receive base pay through the end of 2026, and the company said it would continue health care support through the end of the year.
The jobs left behind look different
Here's the thing if you're building a company now: this isn't really a story about fewer employees. Prince has said he expects Cloudflare to have more total employees in 2027 than at any point in 2026. The question is who those employees are. A company can grow headcount and still make life much harder for people whose jobs are built around coordination, approval chains and internal reporting.
Engineers who can use AI agents well become more valuable in this version of the company. They don't just write code faster. They direct tools, check outputs, design workflows and decide where automation breaks. That is a different skill from merely sitting near a technical team and measuring its progress for someone upstairs. Frankly, a lot of companies have carried too much of that second kind of work because nobody had a better way to keep the machine visible.
Sales also survives Prince's filter, at least for now, because customers still want a person who understands their problem and can be trusted when something goes wrong. That part is easy to overlook in AI restructuring stories. Cloudflare isn't saying all human work is equally exposed. It is saying the safest work either creates the thing being sold or persuades a real customer to buy it.
The uncomfortable part is that this creates a narrower path for many white-collar workers. The old bargain said you could move into operations, management, finance process or internal coordination and build a durable career around making organizations run. Cloudflare is now telling you that some of that glue can be software. Not all of it. Enough of it to cut more than 1,100 roles while adding engineers.
Prince's claim that every company will move this way may be too neat, because every company likes to believe its restructuring is the future arriving early. Still, Cloudflare has put real numbers behind the argument: 20% of staff cut, engineering headcount up 45%, AI usage up more than 600%, and a restructuring charge that runs into nine figures. Those are not vibes. They are the shape of an operating model being rebuilt in public.
The lesson for founders is direct. Don't ask whether AI will replace jobs in the abstract. Look at each role and ask whether it builds, sells or mostly measures. Cloudflare has already answered that question inside its own walls, and more companies will copy the answer if the engineering output keeps rising.
Also read: Arizona is asking for more Colorado River water to feed data centers while facing a 77% cut in its share • Alphabet is betting $180 billion on the idea that owning the whole AI stack is the only way to win • Qualcomm unveils Dragonfly C1000 and signs Meta to reshape how AI gets processed from pocket to data center
