Cloudflare announced its first mass layoff in 16 years on May 8, cutting roughly 1,100 people, or 20 percent of its workforce. The same earnings call reported $639.8 million in quarterly revenue, up 34 percent year over year, the highest in the company’s history. CEO Matthew Prince said the cuts had nothing to do with cost reduction. The reason, in his words, was that internal AI usage had grown by more than 600 percent in three months and the company no longer needed as many people doing the same work.
If you’re job searching in tech right now, “ai layoffs job search” is no longer an abstract framing. Cloudflare made the math explicit, and the framing matters because it tells you what kind of hiring is coming next, what isn’t, and where the openings actually are.
Cloudflare also said it expects to have more employees in 2027 than it did at any point in 2026. The roles being added back, though, are not the ones that were cut. Understanding that gap is most of the job search story for anyone in tech this year.
What Cloudflare actually said on the earnings call
The numbers are worth pulling apart carefully because they set the pattern.
Cloudflare ended the quarter before layoffs with about 5,500 employees. Cutting 1,100 brings headcount to roughly 4,400. CFO Thomas Seifert said on the call that reductions span every team and geography, with one exception: salespeople carrying revenue quotas were excluded. That carve-out matters. It tells you which functions the company still views as scarce and which it views as automatable.
Revenue at $639.8 million was the highest single quarter in company history. Remaining performance obligations, the metric for revenue under contract but not yet delivered, hit $2.5 billion, up 34 percent year over year. The loss for the quarter widened slightly to $62.0 million from $53.2 million in the year-ago period, but as a percentage of revenue the loss shrank. By any normal reading, Cloudflare’s Q1 was strong.
Prince described the internal AI tipping point as November 2025. After that point, he said on the call, “team members that were two, 10, even 100 times more productive than they had been before. It was like going from a manual to an electric screwdriver.” Internal AI usage grew more than 600 percent in three months. The R&D team now uses the company’s own Workers platform for development, including its vibe coding feature, and 100 percent of deployed code is reviewed by autonomous AI agents. Employees across engineering, HR, finance, and marketing run thousands of AI agent sessions a day.
The conclusion Prince drew was direct: highly productive employees using AI require fewer support staff to amplify them. The cut roles were the support layer.
Why this is structurally different from a normal layoff cycle
A traditional tech layoff is usually a cost-cutting move. Revenue weakens, headcount comes down, and when growth returns the company rebuilds roughly the same shape with roughly the same titles. Anyone laid off in cycle one can plausibly target the same job description at a peer company or wait for their old employer to start backfilling.
Cloudflare’s announcement breaks that pattern in two ways.
First, the layoff is happening during record revenue, not despite a downturn. The company’s blog post on the cuts, co-signed by Prince and co-founder Michelle Zatlyn, framed the move as “Cloudflare defining how a world-class, high-growth company operates and creates value in the agentic AI era.” That’s a structural claim, not a financial one. They are not waiting for the next downturn to rebalance. They are doing it now, while in a position of strength.
Second, when Cloudflare hires back, the roles will not match what was cut. Prince said directly that the company will keep hiring people who embrace AI tools because they are “so much more productive than we’d ever seen before.” That means the headcount target for 2027 will be hit with a different role mix. More engineers using AI to do the work of several specialists. Fewer of the support functions that previously absorbed scale. The 1,100 specific roles are unlikely to come back as 1,100 specific roles.
For someone whose title was on the cut list, the search isn’t “find another Cloudflare.” It’s “find work in roles that are still being filled in tech, and reach the people doing the hiring before the role is broadly listed.”
The pattern Cloudflare just joined
TechCrunch noted that Cloudflare’s announcement places it alongside Meta, Microsoft, and Amazon as tech companies reporting revenue growth alongside major layoffs, with both attributed to AI use. Other 2026 announcements have used similar language. As of May, around 92,000 tech jobs have been cut across the sector in the first five months of the year, with a meaningful share of those framed by company leadership as productivity-driven rather than cost-driven.
The phrase being used internally at multiple companies, as reported across Q1 2026 earnings, is some version of “AI-native operating model.” Translated, it means a company structured around the assumption that most knowledge work is done by employees using AI agents, with the headcount expectation set accordingly. The output of one AI-augmented worker is treated as the baseline. The functions whose output was historically additive to that of a senior engineer or a senior PM look like the first to be reorganized away.
This is not coming for every role evenly. Sales quota-carrying positions are notably exempted at Cloudflare and have remained relatively stable across most of the tech layoff wave. Functions where the value is in human relationships, judgment calls under uncertainty, or external-facing accountability are holding up better than functions where the value is in producing artifacts that AI can now produce.
Three categories are absorbing the brunt:
The first is tier-one and tier-two technical support, where ticket triage and resolution can be substantially automated. The second is internal tooling and infrastructure roles where AI agents handle repetitive engineering tasks. The third is content production, data entry, and operations functions where the output is a document, a spreadsheet, or a routine workflow.
If your title was in any of those three areas, the search has to assume the same role won’t be widely advertised again.
What an AI-layoffs job search actually looks like
The candidates who are landing roles in this market aren’t the ones sending the most applications. They’re the ones who figured out early that the listing-based job market lost a lot of its signal during the layoff wave.
Three behaviors are doing the heavy lifting.
The first is targeting companies that are framing themselves as “AI-native” hiring, not companies that already laid off and haven’t backfilled yet. The companies adding headcount in 2026 are those building from scratch around AI-augmented workflows. Their job postings tend to be small in number and large in scope. One posting often covers what would have been three roles in 2023. The candidates getting these positions are reaching the hiring manager directly because the posting itself doesn’t always exist yet when the company starts looking.
The second is repositioning around AI-leveraged output rather than role-specific titles. Recruiters in 2026 don’t want to hire “a content writer.” They want to hire “someone who can run a full content operation using AI tooling, and who can hand the next operator a workflow.” That repositioning has to show up in outreach, not in a resume scan. The resume scan happens after the conversation.
The third is showing up earlier in the hiring process. Companies running an AI-native hiring model often have a shorter listing-to-fill window because they’re hiring against a specific operational need, not a budgeted backfill cycle. By the time a role is on LinkedIn, the hiring manager has often been talking to two or three candidates already. The candidates being chosen are the ones who were in the conversation before the posting went live.
All three of these depend on direct contact. None of them work through a portal application.
The practical playbook
A few things to do right now if Cloudflare’s announcement (or any of the similar Q1 2026 announcements) describes your role.
Identify the 20 to 40 companies you’d actually want to work for that have been framing their hiring as AI-native. Earnings calls and CEO blog posts are the source. Anyone using language like “agentic AI era,” “AI-augmented team,” or “AI-native operating model” is signaling what kind of candidate they’re going to hire. That’s the list.
For each company on that list, identify the function leader you’d report to. Not the CEO, not the head of recruiting. The VP of Engineering, the VP of Marketing, the Director of Customer Success. LinkedIn will give you the names. Glassdoor and company about-us pages fill in the rest.
Reach out with a specific message that demonstrates what AI-augmented output looks like in your role. The bar in 2026 isn’t “I can use ChatGPT.” Everyone says that. The bar is “here’s the workflow I built using AI tools, here’s the throughput it produced, and here’s how I’d apply it to the problem your team is solving.” Send that. Not a resume.
Expect the first conversation to be a working conversation, not a phone screen. Hiring managers running AI-native hiring tend to skip the early gatekeeping rounds and move straight to a 30-minute call about the specific work. If you don’t have a concrete example of the work, the call doesn’t happen.
Follow up on a three-week cadence. Most of the hiring decisions in this market aren’t fast. The companies adding headcount cautiously are adding it cautiously. A candidate who stayed visible without being annoying gets the call when the role finally opens.
What Cloudflare’s announcement doesn’t change
The Cloudflare layoff is unusually clean as data because the company stated explicitly what it was doing and why. Most companies are less direct, but the pattern is the same. AI productivity gains are being used to justify smaller teams, and the teams being built back have a different role mix.
What hasn’t changed is that hiring still happens. Cloudflare itself said it will keep hiring, and that it expects to have more employees in 2027 than at any point in 2026. The same is true across most of the companies that made similar announcements this spring. The volume is down. The shape has changed. The path through the hiring manager has gotten more important relative to the path through a job board, because the job board no longer reliably surfaces the roles being filled.
The roles companies are still hiring for in 2026 are the ones where the value isn’t in producing the artifact, it’s in deciding what artifact to produce, in working with customers and partners, and in operating the AI tools that produce everything else. Those roles tend to be filled by referral or by direct outreach. They are not the roles competing with 400 applicants on Indeed.
If you’re navigating an ai layoffs job search this spring, the most useful thing you can do is stop treating the job board as the primary surface. The companies hiring know who they want to talk to. Your job is to be one of those people before they post the listing.
angld.AI compresses the research-to-outreach pipeline for exactly this kind of search. It identifies the function leader at a target company, surfaces what they’re currently focused on based on recent activity, and drafts a personalized outreach message in about a minute. In a market where the listing-based path is breaking down, the candidates moving through the AI-driven layoff cycle are the ones reaching hiring managers before the role is posted.
