AI News · 8 May 2026
Cloudflare cuts 1,100 jobs and blames the agentic AI era
On 7 May 2026, Cloudflare cut 1,100 jobs and beat its own quarterly numbers in the same press cycle. Revenue grew 34% year over year to $639.8 million. Earnings beat consensus. The stock fell roughly 19% after hours. The CEO memo blamed nothing in particular: not weak demand, not bloat, not poor performance. It blamed agents.
The official story
The memo arrived under both founders’ names, Matthew Prince and Michelle Zatlyn, and it was unusually direct about what it was not. “Today’s actions are not a cost-cutting exercise or an assessment of individuals’ performance,” it said. The framing instead was architectural: “we have to be intentional in how we architect our company for the agentic AI era.” Internally, AI usage at Cloudflare has grown more than 600% in three months. Departing employees keep their full base pay through the end of 2026, healthcare through year-end, and equity vested through 15 August. The company is taking a $140 to $150 million restructuring charge.
AI is driving a fundamental re-platforming of the internet and a paradigm shift in how software is created and consumed. It’s shaping up to be the biggest tailwind we’ve ever seen in Cloudflare’s history.
Prior close $248.85
Regular session closed up 3.3% to $257.04. After the layoff announcement, after-hours trading dropped roughly 19% to $208.20.
Synthesized from price points reported by fxleaders and Cloudflare’s Q1 2026 release. Intraday line is illustrative.
AI was named as the reason in under 8% of US tech layoff announcements across 2025. Through April 2026 it has overtaken every other rationale and now accounts for roughly a quarter of cuts.
Source: Tom’s Hardware analysis of US tech layoff disclosures, compiled April 2026.
The product launch that came first
Twenty-five days before the layoffs, Cloudflare opened Agents Week 2026. From 12 through 17 April, the company shipped what it called the agentic cloud: Sandboxes general availability, Cloudflare Mesh, AI Gateway as a unified inference layer across 14 providers, Browser Run, Artifacts, Cloudflare Email Service, and Agent Memory. The opening blog post pulled no punches: “The Internet wasn’t built for the age of AI. Neither was the cloud.”
Read the timeline cold. Cloudflare’s most ambitious product release of the year ended on 17 April. The 1,100-person reduction landed on 7 May. The internal usage data was already in the Q1 numbers. The reorg followed the launch, not the other way around.

Why the stock fell when peers rose
Earnings cuts dressed in AI language have largely been rewarded in 2026. Microsoft, Meta, and others have trimmed staff with AI as the alibi and watched their stocks rise on the implied margin story. Cloudflare did not get that treatment. The reason is in the guidance, not the headcount.
Q2 revenue guidance came in at $664 to $665 million, a touch under the $666 million Wall Street had penciled in. Full-year revenue was guided to $2.81 billion, just over consensus. Full-year EPS was lifted to $1.19 to $1.20, comfortably above the $1.14 estimate.
That mix tells a specific story. Margins are improving, because removing 1,100 salaries does that mechanically once severance clears. Top-line growth is decelerating, and the soft Q2 number is the shape of that. Investors who bought NET as a high-growth bet noticed the second number more than the first. Shares closed the regular session up 3.3%, then dropped roughly 19% in extended trading to $208.20.
What the memo doesn’t say
The phrase “not a cost-cutting exercise” is a striking one to put in a memo announcing a 20% reduction. The severance is generous and the language is humane, but the disclosure is quiet on a few things a careful reader will look for. It does not name which functions absorbed the cut. It does not specify how much of the $150 million is true severance versus equity acceleration. It does not put the 600% AI usage growth against any productivity benchmark, only against itself three months earlier. And it does not address the basic question of why a company that just shipped its most ambitious agent platform needed to cut a fifth of its staff to operate it.
The founders close with a line that is half promise, half admission: “We’ve asked the team to do this only once… We don’t want to do it again for the foreseeable future.”
What this means for the rest of the industry
Through April 2026, U.S. tech layoffs are running 33% above 2025’s pace, and AI has overtaken every other rationale to become the single most-cited cause, naming roughly a quarter of recent cuts. In 2025 the same figure was under 8%.
Cloudflare’s announcement matters less for the 1,100 jobs than for the template. A profitable, growth-stage infrastructure company shipped a major product, watched its own internal AI usage spike, and decided the implication was a structural reduction in headcount. The agentic AI era, in this telling, is not a productivity multiplier on top of an existing org chart. It is the org chart.
Look out 12 to 24 months and the interesting question isn’t this single restructuring. It’s how the template generalises. Engineering, support, content review, and ops are the obvious early candidates. Finance, legal, and sales ops follow when hallucination rates drop below the human error rate they’re replacing. By mid-2027, the AI plan and the headcount plan are the same paragraph in every operating plan, not two separate documents. The companies that rebuild around what agents do well, orchestration, throughput, after-hours coverage, spend the next two years compounding. The ones still bolting agents onto an unchanged org chart spend it cutting twice.