On February 27, 2026, Block (formerly Square) CEO Jack Dorsey sent a memo to all staff. "The intelligence tools we're building, combined with smaller and flatter teams, are fundamentally changing what it means to build and run a company." And on that same day, he handed pink slips to more than 4,000 employees — 40% of the company.
Wearing a hat that said "LOVE" as he delivered the news, Dorsey added: "Our business is solid. Gross profit continues to grow." The message: this wasn't about financial trouble — it was about the AI-driven future.
The stock shot up 22% that day. But former data scientist Naoko Takeda posted on LinkedIn: "For the past year, AI has been shoved down everyone's throats. Forcing people to use tools that eliminate the very jobs their livelihoods depend on — that's dystopian."
So How Did This Happen?
Block runs Square, Cash App, and Afterpay. It had 3,835 employees in 2019, then exploded during the pandemic to over 12,000 by late 2022. Dorsey's layoffs brought the headcount back below 6,000.
Dorsey's logic was explicit. Five weeks after the layoffs, he co-published an essay with Sequoia Capital's Roelof Botha called "From Hierarchy to Intelligence." The core argument: corporate hierarchy is an information routing protocol that dates back 2,000 years to the Roman army — and AI can now replace the coordination function that middle managers used to serve.
Dorsey put forward the formula "100 people + AI = 1,000 people" and set a guidance target that the remaining employees needed to 2.6x their productivity within one year.
But seven current and former employees interviewed by The Guardian told a very different story.
"95% of AI-generated code doesn't meet company standards and needs human rework. What's technically possible and what a CEO preaches from his own interpretation are completely different things."
Employees described AI use shifting from "recommended" to "mandatory." Token consumption was tracked alongside usage metrics, and AI proficiency was added to performance reviews. One engineer said: "It was clear that if you weren't using AI, you were a layoff target."
What made it worse was that employees were put in the position of building and training the very AI that would replace them. They were required to submit weekly reports on which tasks could be automated — and that data ultimately became the basis for layoff decisions.
What Changes?
Here's the thing — about a year earlier, another company ran almost the same experiment. That company was Klarna. And it went in the exact opposite direction.
| Block (Dorsey) | Klarna (Siemiatkowski) | |
|---|---|---|
| Scale of cuts | 4,000 people (40%) | ~1,500 people (30%) |
| Official reason | "AI changed how we work" | "AI chatbot replaced 700 support agents" |
| Stock reaction | +22% (day of layoffs) | Initial bump, then $40B market cap drop |
| One year later | Still playing out (Apr 2026) | Re-hired humans, CEO admits mistake |
| CEO quote | "Most companies will make the same decision within a year" | "Cost was the dominant metric. Low quality was the result" |
| Employee reaction | Hundreds of thumbs-down, tomato, and clown emojis on internal Slack | Data scientists quit + LinkedIn exposés |
| Hidden context | 3x pandemic over-hiring, $68M company party 5 months prior | BNPL market shrinkage, IPO pressure |
The common thread between the two is clear: both used an AI narrative to cover for management failures.
Om Malik called it "Narrative Substitution." Reframe a major operational mistake as an AI transformation, and investors and the public start viewing the situation through a completely different lens.
Mizuho analyst Dan Dolev also concluded that "the vast majority of these layoffs aren't about AI," and even OpenAI's Sam Altman has called this kind of pattern "AI washing."
That said, Block does differ from Klarna in one important way. Block isn't just swapping CS staff for AI — it's declared a full redesign of its organizational structure. The "From Hierarchy to Intelligence" essay lays out three roles — IC (individual contributor), DRI (directly responsible individual), and player-coach — that Dorsey says are already running internally.
The real question isn't "Can AI replace people?" It's "Is leadership using AI as a genuine strategy, or as an excuse?"
The Bottom Line: What Should Your Organization Do?
The lesson from Block and Klarna's experiments is pretty clear. AI adoption itself isn't the problem — the danger starts the moment you approach it through a "replacement" frame.
- Start with "augmentation," not "replacement"
Block's mistake was telling employees to build "tools that replace you." That destroys morale and organizational trust. Instead, ask: "What can you do 10x better with this tool?" That's the direction Klarna's hybrid model eventually moved toward. - Before announcing "AI-driven" layoffs, verify the numbers actually came from AI
Block's operating income per employee was the lowest in fintech ($167K vs. Adyen's $281K). That's an operational efficiency problem, not an AI story. If AI genuinely boosted productivity, measure exactly which processes improved and by how much. "100 people + AI = 1,000 people" is a bumper sticker, not a business plan. - Roll out AI tools gradually, with an off-ramp
Klarna cut 700 people at once and had to rehire a year later. Block cut 4,000 in a single day. IBM research shows 75% of AI projects miss their promised ROI. Increase AI's share of work in 10–20% increments and check quality metrics as you go. - Treat employees' AI use as something to support, not surveil
Block tracked token consumption and mandated AI use. The result? Internal reaction: "People are fed up with AI." Accenture also drew criticism for announcing it would factor AI usage monitoring into promotion decisions. Don't force it — find workflows where AI genuinely speeds things up, and show that to your team. - Separate the AI narrative from management accountability
Dorsey actually admitted the over-hiring on X: "During COVID, I made the mistake of treating Square and Cash App as two separate companies." That's a management failure, not an AI story. When you conflate the two, your organization gets confused and the market misses what's actually going on.
Want to go deeper on the Klarna story?
The Klarna Lesson — Why They're Re-Hiring After Replacing 700 People with AI covers the hybrid CS model transition in detail, with a practical guide.
