Cloudflare Lays Off 20% Despite Beat-and-Raise, Stock Falls 18%: First Time AI Usage 600% Surge Cited as Documented Justification

On May 7, 2026, Cloudflare announced layoffs of approximately 1,100 employees, about 20% of its workforce, alongside its earnings release. First-quarter revenue grew 34% year-over-year to $639.8 million, and adjusted EPS beat consensus. Yet the stock fell 14% to 18% in after-hours trading.

CEO Matthew Prince wrote in an internal memo that "Cloudflare's usage of AI has increased by more than 600% in the last three months alone," and explained the layoffs as "not a cost-cutting exercise." Reading the 8-K filed with the SEC the same day together with the company-wide memo, Cloudflare becomes the first major IT firm to put its rationale for AI-driven workforce replacement into formal documents using a concrete, numerical metric drawn from internal usage logs.

Past layoffs were justified with abstract phrases like "organizational restructuring," "efficiency gains," or "performance bar resets." Today's announcement marks the moment that justification language evolved from abstraction to numerical evidence. Let's walk through it in order.

What Cloudflare actually does

Cloudflare is a U.S. company that provides infrastructure to speed up websites and protect them from cyberattacks. Headquartered in San Francisco, founded in 2010, with about 5,150 employees today. Its services run behind a wide range of news sites, e-commerce platforms, and SaaS products globally.

The core offerings fall into three buckets. First is the CDN (content delivery network): servers around the world deliver web pages from a location close to each user, making sites load faster. Second is the WAF (web application firewall), which blocks DDoS attacks and unauthorized access. Third is Workers, a serverless code execution environment that runs JavaScript and WebAssembly across the company's entire data center fleet.

Many international companies use Cloudflare for site acceleration, secure delivery to overseas users, and API protection. The scale is large enough that an outage at Cloudflare can cascade across major sites worldwide.

What the 8-K and the internal memo say

The layoff announcement came as a two-track release: the 8-K (the form U.S.-listed companies use to disclose material events), filed alongside the earnings release, and a company-wide memo from CEO Matthew Prince to all employees.

The 8-K (filed May 7 at 4:21 PM ET, signed by Chief Legal Officer Alissa Starzak) states that Cloudflare will book restructuring charges of $140 million to $150 million ($105M–$110M in cash plus $35M–$40M in non-cash stock-based charges), with the majority falling in the second quarter and execution substantially complete by the end of Q3 2026. Prince is quoted in the filing: "With AI and agents now core parts of our workforce, the way we work at Cloudflare has fundamentally changed."

The internal memo goes further. Co-signed by Prince and co-founder Michelle Zatlyn, it reads:

The memo explicitly states the layoffs are "not a cost-cutting exercise" and "not an assessment of individuals' performance," framing them instead as a reorganization to operate in what the company calls the "agentic AI era."

Severance terms are relatively generous by tech industry standards: full base pay through the end of 2026, continued U.S. healthcare through year-end, accelerated equity vesting through August 15, and waived one-year cliff requirements with prorated vesting through August.

On the earnings call, Prince said: "This wasn't an easy decision, but it's the right decision," and added that "there are roles at the company that just aren't the roles that we need for the future." In the press release, he also wrote that "AI is driving a fundamental re-platforming of the internet — the biggest tailwind we've ever seen in Cloudflare's history."

Layoff justification language has moved from abstraction to numbers

What separates this announcement from earlier ones is the texture of the rationale. Since the start of 2026, four AI-cited mass layoffs have unfolded. Reading their "why we're cutting" statements in order makes the linguistic progression visible.

March 11: Atlassian, 1,600 layoffs (~10% of workforce). CEO Mike Cannon-Brookes wrote that "AI changes the mix of skills and the number of roles we need." Critics noted the contradiction with his statement five months earlier that "AI lets us hire more engineers," but the framing itself stays within the abstraction of "organizational restructuring."

Late March: Block, 4,000 layoffs (~40% of workforce). Jack Dorsey's company cited "efficiency" and "AI utilization." The stock rose 23% on the announcement, prompting "AI washing" allegations. The rationale, again, is the abstraction of "efficiency."

March 21: Meta, considering up to 16,000 cuts. Meta tied the move to "funding $135 billion in AI investment by trimming personnel costs." Half-numerical, half-abstract.

March 27: Epic Games, 1,000+ layoffs (~20% of workforce). Epic's reasoning was performance-driven: "Fortnite playtime fell 24%." AI was not the headline.

May 7: Cloudflare, 1,100 layoffs (~20% of workforce). Here, "internal AI usage rose 600% in three months" emerges for the first time as the explicit, document-of-record rationale. An internal system metric — externally hard to verify but precisely measurable inside the company — has been written into the justification for layoffs.

From Atlassian's "restructuring" to Block's "efficiency" to Meta's "funding source" to Cloudflare's "600% AI use," the language steadily moves from abstraction toward numerical evidence. The crucial part is that Cloudflare's number is not financial performance — it is internal usage. The same logic transfers cleanly to any other company that has begun heavy internal AI use.

Why a beat-and-raise quarter sent the stock down

Cloudflare's Q1 was a beat by the numbers. Revenue of $639.8 million topped the $620.83 million consensus, and adjusted EPS of $0.25 beat the $0.23 estimate. The stock still fell 14–18% in after-hours trading.

Three triggers explain the drop.

First, a slight Q2 guidance miss. Cloudflare guided next-quarter revenue to $664–$665 million, against consensus of $666.1 million. The dollar gap is small — a few million on a $665M base — but the market read it as the leading edge of decelerating growth.

Second, the concentration of restructuring charges. The bulk of the $140–$150 million layoff cost lands in Q2, compressing operating income that quarter. The company gets leaner only after a quarter of compressed margins.

Third, skepticism about whether a 20%-smaller organization can keep growing. Prince's argument is that AI fills the gap. The market will wait six to twelve months for that to show up in numbers. Between the CEO's thesis and the eventual data sits the share price, oscillating.

For full-year 2026, Cloudflare guided to $2.805–$2.813 billion, narrowly above the $2.8 billion consensus. "Strong Q1, soft Q2, slightly raised full-year" — the market chose to weight the Q2 structure most heavily.

What this means for enterprise customers and how to read it

Enterprise use of Cloudflare typically spans three layers: CDN (acceleration), WAF (defense), and Workers (code execution). Whether the layoffs degrade service quality depends on how support and engineering teams are reshaped. Cloudflare has historically been respected for incident response speed; procurement and platform teams should monitor support SLAs and incident handling closely in the months ahead.

The other thread to follow is the spread of "using internal AI logs as headcount-planning evidence" as an operating model. Cloudflare's argument transfers easily to any company that runs AI agents heavily inside its workflow. Aggregate the daily AI agent session count from Slack, IDE assistants, and internal tooling, divide it by some productivity factor, and the equation "we use this much AI, therefore we need this many fewer people" starts to write itself.

For consulting firms, SIers, and SaaS vendors, the same logic is likely to enter headcount planning conversations. Data already exists showing AI has changed software development productivity, and "usage volume as the reason" will increasingly appear on the table for workforce reviews.

The day "cutting people by the numbers" became standard

The notable thing about today's announcement is less the scale than the shift in language.

Past mass layoffs were justified with phrases like "deteriorating performance," "organizational restructuring," "efficiency" — all hard to verify from outside. Cloudflare put a concrete number from its own internal systems — "AI usage up 600% in three months" — into the document of record. The logic of replacing humans with AI has, for the first time, made it into a public corporate filing with numerical backing.

Prince and Zatlyn repeatedly stress this is "not a cost-cutting exercise." The market's reply was a 14–18% drop. The CEO's frame and the frames used by shareholders, the labor market, and regulators are likely to diverge for a while.

By the time 2026 ends, today may be remembered as the day the logic of cutting people for AI shifted from abstraction to numbers. Within two months — from Atlassian's "restructuring" to today — the language has traveled this far. The next company to make this argument no longer needs a footnote: they can simply write "our AI usage rose X%." That template was finalized today.

References

• ▸ Bloomberg - Cloudflare to Cut 1,100 Jobs as It Shifts to AI-First Operating Model (May 7, 2026)
• ▸ CNBC - Cloudflare stock sinks 18% after earnings as company cuts 1,100 employees due to AI changes (May 7, 2026)
• ▸ SEC 8-K - Cloudflare, Inc. Reports Material Event (Filed May 7, 2026, 4:21 PM)
• ▸ AOL/Fortune - Read the memo: Cloudflare is laying off 1,100 employees to prepare for 'the agentic AI era' (May 7, 2026)
• ▸ SiliconANGLE - Cloudflare beats on earnings, but 20% AI-driven layoffs and weak guidance send shares down (May 7, 2026)
• ▸ The Motley Fool - Cloudflare NET Q1 2026 Earnings Call Transcript (May 7, 2026)