Block Fired 40% of its People. Wall St. Cheered. It’s Just the Start.

Jack Dorsey was wearing a hat that said “LOVE” when he fired nearly half his company.

During the all-hands videoconference where he explained the decision to cut more than 4,000 employees – roughly 40% of Block’s global workforce – dozens of thumbs-down emoji cascaded down the screen. One employee asked whether the hat was really the right fashion choice for the occasion.

Dorsey acknowledged the tension directly. “I’d rather it feel awkward and human than efficient and cold,” he wrote in his note to employees. That sentence captures the entire story in ten words.

Block’s stock jumped more than 26% in after-hours trading.

Read that again. A company just fired four in ten of its own people, and investors delivered one of the best single-session reactions in recent memory. That is not sentiment. That is the market declaring, in the clearest possible terms, what the new rules are.

What Is Driving AI Layoffs in 2026?

Dorsey did not hide behind consultant-speak. In his note to employees and his shareholder letter, he made the logic explicit: “Intelligence tools have changed what it means to build and run a company. We’re already seeing it internally. A significantly smaller team, using the tools we’re building, can do more and do it better. And intelligence tool capabilities are compounding faster every week.”

He then said something no major CEO has said this directly before:

“I think most companies are late. Within the next year, I believe the majority of companies will reach the same conclusion and make similar structural changes. I’d rather get there honestly and on our own terms than be forced into it reactively.” – Jack Dorsey

He also explained his timing logic with unusual candor: “I had two options: cut gradually over months or years as this shift plays out, or be honest about where we are and act on it now. I chose the latter. Repeated rounds of cuts are destructive to morale, to focus, and to the trust that customers and shareholders place in our ability to lead.” That framing explains the market reaction as much as the efficiency narrative does. Investors are not just pricing in lower labor costs. They are pricing in a management team willing to be decisive.

This is not a cost-cutting story dressed in AI language. Block had a strong 2025. Gross profit growth more than doubled from Q1 to Q4. The company surpassed the Rule of 40 in Q4, reignited Cash App network growth, drove record new volume added for Square, and scaled lending products at strong returns. Dorsey said the financial results are “just beginning to reflect the product development velocity improvements” the company drove last year.

They cut 4,000 people from a position of strength. That is the part that should get every CEO’s attention.

For those who lost their jobs, Dorsey announced: 20 weeks of salary, plus one additional week per year of tenure, equity vested through the end of May, six months of healthcare coverage, corporate devices, and $5,000 in transition support. It is a more substantial package than most layoff announcements include, and Dorsey led with it deliberately. “I want you to know that before anything else,” he wrote. The severance terms will become a benchmark that other boards will be measured against.

Why AI Job Cuts Are Spreading Beyond Tech

The conventional narrative is that software engineers are the first to feel AI displacement because AI can write code. That is correct and dangerously incomplete.

Software is simply where the decoupling becomes visible first, because the inputs and outputs are most legible to AI systems. But the economic logic applies with equal force to customer service, financial analysis, marketing operations, legal research, HR administration, and procurement. The question is not whether AI job displacement spreads beyond software. The question is how fast.

Block is not alone in arriving at this conclusion. Klarna reduced its workforce from 5,527 employees in 2022 to fewer than 3,000 today. Its AI assistant now handles the work of 853 full-time customer service agents. Average revenue per employee rose 73% year over year. Salesforce has said AI handles 50% of its workload. Amazon cited “fewer layers.” Pinterest, CrowdStrike, and Chegg all named AI explicitly in recent reductions.

Anton Korinek, an economist studying transformative AI at the University of Virginia, put it directly: “This may be the beginning of a new trend where white-collar jobs become threatened more seriously by AI. Once a few companies start the trend, competitive forces may induce others to follow suit.”

Dorsey’s 12-month prediction is not a distant forecast. That is the next business planning cycle.

The Three-Layer Workforce: Where Does Every Job Actually Belong?

Here is what the layoff headlines are not explaining. The enterprise workforce is not disappearing. It is reorganizing into three distinct layers, and understanding which layer each function belongs in is the most important strategic question any executive can answer right now.

The Autonomous Layer is a fully AI-powered foundation of digital agents executing work that should have been automated years ago. Repetitive analysis, rule-based decisions, data processing, content generation at scale, customer query triage, standard contract review. All of it handled continuously, without coordination costs, attention limits, or the error rates of human teams performing the same tasks.

The Judgment Layer is an exclusively human domain for decisions that carry consequences too significant, too contextual, or too ethically complex to delegate. Business strategy, governance, creative vision, stakeholder relationships, brand character, risk accountability. This layer does not shrink under AI pressure. It becomes more valuable, because the humans in it are finally freed from the operational noise that consumed most of their time.

The messy middle is where the real instability lives: a Human-Plus-Machine layer where the boundaries between human and digital labor are neither clean nor stable. It takes two forms that will coexist uneasily for the next decade. In one version, humans manage AI agents, orchestrating and supervising fleets of autonomous systems. In the other, AI systems route tasks, coordinate human work, and optimize output in ways that reduce workers to components in a larger automated whole.

Economic pressure flows in one direction: toward whichever configuration maximizes output per dollar. The middle layer stays hybrid only as long as human judgment adds enough value to justify the premium. When it does not, the layer collapses toward automation.

What the Klarna Reversal Actually Proves About AI and Jobs

Last year, Klarna became the cautionary data point. After its AI-first transformation, customer service quality deteriorated. CEO Sebastian Siemiatkowski publicly admitted the company had gone too far: “Cost unfortunately seems to have been a too predominant evaluation factor.” Klarna began rehiring.

Critics pointed to this as evidence that AI replacement has hard limits. They are right that it has limits. They are wrong about what those limits mean.

Klarna did not fail because AI was inadequate. It failed because the company deployed AI in a context where the Judgment Layer was non-negotiable. A customer disputing a fraudulent charge on a financial product does not want to explain their situation to a system that cannot feel urgency, cannot handle nuance, and defaults to scripts. That is Judgment Layer work wearing an Autonomous Layer uniform. Klarna learned that at the cost of brand trust and service quality.

Block is making a different calculation. Dorsey is not claiming he has automated judgment. He is claiming a smaller, more capable team armed with AI can do everything his 10,000-person organization was doing. Critically, he framed this as a mission imperative, not just an efficiency play. Block “serves millions of customers,” he wrote, including “small businesses that rely on us to get paid, to manage their money, to access capital” and “individuals navigating a financial landscape that’s changing fast.” His argument is that getting leaner makes Block faster at serving the people most vulnerable to the same AI disruption that just eliminated their jobs. Whether that logic holds is a question the next two years will answer.

The Math Every CEO Is Now Running

“Intelligence tool capabilities are compounding faster every week.” That sentence, from Dorsey’s shareholder letter, is the one that should keep every executive up at night.

Model that compounding rate against the trajectory of human labor costs, benefits inflation, and organizational complexity, and you arrive at exactly the conclusion investors reached when they pushed Block’s stock up 26% in a single session. The math favors digital labor. Not eventually. Now. And the margin widens every month.

For every function in your enterprise that belongs in the Autonomous Layer, that math is already your competitive reality. Delay is not a neutral choice. Delay is a decision to cede margin to competitors who move faster.

Three Questions Every Executive Needs to Answer

Dorsey’s announcement is a forcing function for every boardroom and C-suite. The question is no longer whether to restructure around this model. The question is whether it happens on your terms or under duress.

First, map every major function against the three-layer framework. What is genuinely Judgment Layer work requiring human accountability and contextual wisdom? What is already Autonomous Layer ready? What sits in the messy middle requiring careful design before you automate it?

Second, audit your software engineering and operational teams specifically. Software is where this shift is most advanced and most legible. If you have not yet assessed what portion of your development capacity could be handled by AI agents alongside a smaller team of senior engineers, your assumptions about your cost structure are already wrong.

Third, ask the governance question directly: do your compensation committees and risk frameworks account for a world where headcount is no longer a proxy for capability? Many governance structures still reward scale. The new economics punish it.

The Bigger Signal Hidden Inside the Announcement

The workforce story is the headline. The business model story is the one leaders are missing.

In his note to employees, Dorsey described a future where Block’s customers “can build their own features directly, composed of our capabilities and served through our interfaces.” Read that carefully. Block is not just cutting headcount to reduce costs. It is repositioning itself as an AI infrastructure layer that small businesses assemble into their own products. That is a fundamental shift in what Block is: from a payments and financial services company to a composable platform that businesses build on top of.

This is what platform inversion looks like in practice. The customer relationship no longer runs through Block’s products. It runs through the customer’s own AI-assembled interface, powered by Block’s underlying capabilities. For Block, this is a defensible moat strategy. For every traditional business that has not made a similar move, it is a competitive threat that does not show up in this quarter’s numbers but will define the next five years.

The Cost the Numbers Cannot Capture

More than 4,000 people at Block are losing their jobs. Those are people with mortgages, families, and careers built over years. The Judgment Layer is real, but it employs far fewer people than the Autonomous Layer absorbs. The messy middle requires new skills that current workers may not have and retraining pipelines that do not yet exist at scale.

Dorsey at least acknowledged this dimension directly. Block exists to help small businesses and individuals “navigate a financial landscape that’s changing fast,” he wrote. The darkest irony in the announcement is that the same AI transformation clearing Block’s books is accelerating the financial instability of the very customers Block says it was built to serve.

What makes his announcement different from most CEO techno-optimism is the absence of reassurance. He is not telling you the disruption is painless or distant. He is telling you it is here, it is accelerating, and the companies that do not plan for it will be forced into it without preparation. He would rather it feel awkward and human than efficient and cold. But either way, it is happening.

He is right about that.

The question worth sitting with is not whether the three-layer workforce replaces the old model. It will. The question is who designs that transition, who bears its costs, and who captures its gains. That is not an engineering problem. That is a governance and policy problem. And it needs to move as fast as the models.