Cloudflare Cut 1,100 Jobs. AI Usage Surged 600%. One of These Things Is Related

May 7, 2026. Cloudflare reports Q1: $640M revenue, 34% YoY growth, 25 cents EPS (beat 23 cents expected). Great numbers. Same day: 1,100 people gone, 20% of workforce. Stock down 18% after hours.

CEO Matthew Prince calls it an “agentic AI-first operating model.” Points at 600% internal AI usage spike.

Reid Hoffman posted within hours: companies using AI to dress up ordinary cuts.

He’s right. But the more useful question isn’t “is this AI washing?” It’s “what does the 600% number actually tell us?”

Here’s what I keep thinking about.

Cloudflare lost $22.9M net income in Q1. Not profitable. The “AI made us efficient” story works for headlines. Cost-cutting works for balance sheets.

The Gartner Study Nobody’s Reading

Same day Cloudflare cut, Gartner dropped research nobody linked. 80% of companies deploying autonomous AI cut staff. Fine. Here’s the part that should concern every exec citing AI as their layoff rationale: workforce reduction rates were identical between companies with strong AI returns and companies with weak or negative returns. Their words: “Workforce reductions may create budget room, but they do not create return.”

The cuts happened. AI adoption happened. Correlation, not causation.

Both can be true without one explaining the other.

The framing serves a purpose. “We restructured around AI” sounds like forward thinking. “We overhired during the pandemic and the org got fat” sounds like a mistake. One plays well on the earnings call.

The other gets you replaced.

The 600% Is the Interesting Part

Forget the layoffs for a second. 600% AI usage growth in three months.

Thousands of agent sessions daily across engineering, HR, finance, marketing. This is what enterprise AI looks like running for real, not in a PowerPoint.

Ask yourself: what are they running? How do they measure? What broke first and how did they fix it? Nobody’s writing about this because it’s not dramatic. The drama is 1,100 people.

The actual story is what happens when a enterprise runs thousands of AI sessions a day and realizes they need fewer people to manage the output.

For small shops, this is an opening.

You can run those same workflows. You don’t need the org chart to do it. A lean team with good agentic tooling punches way above what it did two years ago. The gap between a three-person shop running AI well and a 50-person team arguing in Jira is not a gap anymore. It’s a chasm.

Side note: Cloudflare had planned to hire 1,111 interns before cutting 1,100 full-time staff. That’s not an AI story. That’s just a bad headline.

The Tech Layoffs Are Bigger Than AI

PayPal cut 20% over three years.

Coinbase 14%. Block 40%. 92,000 tech jobs gone in 2026 so far. Cloudflare got the coverage since “AI” was in the press release. The others just said “restructuring.”

If you’re in tech right now watching this, here’s my take: the layoffs aren’t about you being replaceable. They’re about companies using a convenient narrative. You are not less valuable than you were two years ago. The labor market is moving weird and companies are opportunistic about it.

The skill that compounds isn’t any specific task AI can do faster.

It’s knowing how to build the workflow, measure the output, and own the outcome. Agents need handlers. Become the handler.

What You Do With This

Don’t use AI as a cover for bad decisions. If you’re cutting, say you’re cutting. If AI drove the cut, show the data. “Usage went up 600%” isn’t the same as “we got 600% more productive.”

If you’re running lean already, this is your moment. Big teams are shrinking. You can move faster, adopt faster, iterate faster. The companies figuring out agentic workflows at scale first won’t need to hire.

They’ll just stop posting job reqs and pretend that’s normal.

Cloudflare’s numbers are fine. They’ll recover. The 1,100 people who lost their jobs will too. The market keeps moving.