Samsung Cuts 800 US Workers, Discovers AI Profits Don't Need People

Samsung Electronics America has eliminated over 800 U.S. positions, with 739 cuts at its Englewood Cliffs, New Jersey headquarters and approximately 100 more in Plano, Texas. The timing contains a particular cruelty: these employees had moved into new offices in Englewood Cliffs less than a year ago, in September 2025, with the kind of corporate pageantry that typically signals stability rather than the preamble to a bloodbath.

The company's official language for this restructuring is characteristically bloodless. Samsung said it may make "changes in our workforce structure" for "employees who are unable to relocate, or certain functions that are optimized to ensure our roles align to key business priorities." Translation: move to Texas or exit through the side door. Either way, the headcount problem solves itself.

What makes this layoff legible within the broader tech economy is not Samsung's need to cut costs—it's the mathematical nakedness of what they're cutting for. Samsung's semiconductor division has just posted a 19-fold profit surge, driven almost entirely by AI chip demand. In July alone, the division generated 53.7 trillion won in operating profits. Those are generational numbers. The company is not restructuring because it lacks resources. It is restructuring because those resources now belong to silicon, not staff.

The contrast within Samsung is instructive. While semiconductors are generating bonanza profits, the mobile division is expected to post its first-ever loss as it gets compressed between Apple at the premium end and Chinese competitors below. Rather than sustain an inefficient division, Samsung is making the economically rational choice: place all bets on the division that's actually printing money. The workers in between become a problem to be optimized away.

This is now the dominant pattern across technology. Microsoft, Amazon, and Meta have all trimmed workforces while simultaneously increasing investment in AI infrastructure and data centers. The message is consistent: future growth comes from silicon and scale, not from dispersed human problem-solving. If you're not directly enabling that bet, you're a line item to be eliminated.

What makes the situation genuinely dark is the economic contradiction now baked into AI economics. The entire valuation thesis for AI companies depends on a future where artificial intelligence generates extraordinary returns. But those returns are increasingly predicated on replacing the labor force that consumes the products generating those returns. Samsung's semiconductor division is profitable because tech companies are spending like drunken sailors on AI chips. Those tech companies are spending aggressively partly because they believe AI will let them operate with fewer expensive humans. The consumer base that validates all of this is simultaneously being laid off.

It is a recursion that would concern a serious analyst: profitability increasingly means profitability achieved through workforce reduction. The economic model is not scaling labor efficiently. It is eliminating labor as a cost category. When that logic extends across the technology sector—which it clearly is—you have created an economy where shareholder value is increasingly generated by destroying the purchasing power of the people supposed to use the products.

Samsung's move to Texas is part of a larger industrial migration. Lower taxes, reduced regulatory burden, and a business-friendly environment clearly matter. But the real story is that Samsung's most profitable division is now located anywhere there's semiconductor manufacturing capacity and cheap electricity. Workers, by contrast, were a friction point. The Englewood Cliffs office was a real estate decision. The layoffs are a profitability decision. Texas got the former. The latter went to the spreadsheet.