This story landed in every inbox in America last week as proof that AI is coming for white-collar jobs. Bloomberg's own headline called it "AI-washing." Neither frame is the most useful one. Here is what actually changed, and what it means if you are paying attention to your own situation.
Block Cut 4,000 Jobs and Said AI Did It. What That Actually Means.
Block Inc., the company behind Square and Cash App, cut 4,000 jobs last month, nearly half its global workforce, and co-founder Jack Dorsey said AI made it possible. The business was profitable. Gross profit was up 24% year on year. The stock went up 24% the same day the announcement was made. We read the original shareholder letter, the Q4 earnings call transcript (the formal investor briefing that accompanies quarterly results), Bloomberg's reporting, and the Oxford Economics study (an independent economic research report released in January) cited in subsequent coverage, specifically to understand whether this is a new kind of event or a familiar one in unfamiliar clothes.
It is probably both. And that distinction matters more than either the "AI job apocalypse has begun" reading or the "this is just COVID overhiring being unwound" reading that critics offered. What changed last month is not that AI took 4,000 specific jobs. What changed is that a public company eliminated nearly half its workforce while the business was growing, and the market rewarded it. That ratio is new. For anyone in white-collar work, it changes which questions are worth asking about your own situation, not next year, but now.
Dorsey's case for the cuts was specific. AI tools had made the same work achievable with a significantly smaller team. "Something happened in December," he told analysts, "where the models just got an order of magnitude more capable." He named customer support, document processing, and software development as the areas most transformed. He predicted most companies would reach the same conclusion within a year.
Serious critics pushed back immediately. Block had nearly tripled its headcount between 2019 and 2023, hiring during the pandemic for longer than its rivals. An Oxford Economics report released in January found that many layoffs CEOs describe as AI-related are actually correction for prior overhiring. Bloomberg called the announcement "AI-washing," using AI as a convenient label for decisions that had other drivers. Dorsey acknowledged the overhiring on X but disputed that it explained the scale of the cuts.
Here is the honest editorial position: both things are probably true, in proportions that are genuinely hard to separate from the outside. The overhiring correction is real. The AI capability shift Dorsey named is also real. The December 2025 model improvements are independently documented. What the debate between these two framings misses is the question that actually matters: what happens when the next company, one that did not overhire, reaches the same conclusion? This morning, Fortune reported that Meta may be next.
The jobs Dorsey named as most transformed are specific categories of white-collar knowledge work. They are not every job. They are also not an abstract category.
This is the situation as of March 2026. Dorsey's prediction about other companies following suit, and the pace of AI capability change, are both evolving. Worth revisiting in three to six months. The story will look different.
Limitation:
This is the situation as of March 2026. Dorsey's prediction about other companies following suit, and the pace of AI capability change, are both evolving. Worth revisiting in three to six months. The story will look different.
In conversation:
A big fintech company cut 40% of its staff last month, said AI made it possible, and the stock went up the same day, which is a different kind of signal than what most of the headlines were actually saying.
Questions people ask
Was this really about AI, or was Dorsey using AI as cover for cutting pandemic overhiring?
Probably both, in proportions that are genuinely difficult to separate. Block did overhire significantly. Its workforce tripled between 2019 and 2023. But Dorsey named December 2025 specifically as the month AI capability crossed a threshold, and the model improvements during that period are documented independently. The more useful question: even if the trigger was overhiring, does the AI shift mean the correction goes deeper than it otherwise would have?
Which jobs are actually most at risk from AI right now?
The categories Dorsey named (customer support, document processing, certain software development tasks) match what independent analysts are tracking across the industry. Roles requiring complex judgment, physical presence, or sustained relationship management are less directly exposed. The line between task-based and judgment-based work is the one worth understanding in your own role. The Pragmatist would frame it this way. The question is not which jobs, but which tasks within your job.
If my employer starts making similar moves, what should I be doing right now?
Understanding which parts of your role are task-based versus judgment-based is genuinely useful preparation, not as a defensive strategy, but as a way of understanding where your actual value sits. The Bigger Picture post "Is My Job Actually Safe?" addresses this directly. Members can also bring a specific version of this question to Ask Kevin.