Record Spending on Artificial Intelligence - But 4,800 People Lose Their Jobs

There is an equation that does not seem to add up at first glance, and that is precisely where its bitter logic lies. Microsoft has announced the layoff of approximately 4,800 employees, about 2.1 percent of its total workforce. At the very same time, the company is spending more money on artificial intelligence than ever before. The two belong together, even if they appear to contradict each other. Companies across the technology sector are increasing their investments in artificial intelligence infrastructure while using that same technology to streamline operations and eliminate part of their workforce. The machines they invest in are replacing the people they no longer need.

The numbers are staggering. The world's largest technology companies are expected to spend a record more than $700 billion on artificial intelligence this year. Those enormous investments are forcing companies to find ways to recover their costs and offset the rising expense of deploying AI. Alongside Microsoft, both Amazon and Meta have laid off thousands of employees this year. This is not an isolated case. It is a pattern, and it is a pattern that almost no one is willing to name openly. The price of this expensive future is paid first by the people who lose their jobs today. Microsoft announced the layoffs on Monday after its stock had fallen nearly 23 percent during the first half of 2026, its worst performance since 2022. Earlier this year, the company had already offered voluntary severance packages to roughly 9,000 U.S. employees, about 7 percent of its American workforce. Microsoft has traditionally reduced its headcount toward the end of its fiscal year, which closes in June, in order to optimize the budget for the next financial period. The company schedules layoffs the same way it schedules items on a balance sheet, carefully timed to coincide with the end of the fiscal year.

The driving force behind all of this is demand for artificial intelligence, which continues to fuel the growth of Azure, Microsoft's cloud platform that, until April, was the only service through which customers could access OpenAI's models. Yet the cost of building the data centers that make those services possible is placing tremendous pressure on the company's cash flow. In April, Microsoft projected Azure sales above Wall Street expectations, while simultaneously announcing plans to invest approximately $190 billion in infrastructure during 2026, far more than analysts had anticipated. It is a race in which spending is growing faster than revenue, and in such a race people become the variable that can be adjusted whenever the numbers no longer work. Even businesses that once seemed untouchable are now under pressure. Microsoft's Xbox division illustrates that reality. Last month, Xbox executive Asha Sharma said the business needed a reset, explaining that operating margins had fallen to just 3 percent, forcing the company to restructure and even consider acquisitions and mergers. Without Activision Blizzard King, she wrote in an open letter to employees, Microsoft had invested more than $20 billion over the past five years in content, platforms, and hardware subsidies, yet annual revenue still declined by nearly half a billion dollars. Twenty billion dollars invested, half a billion dollars less in revenue. It is an equation that captures the entire dilemma facing this industry.

This is where the story reaches beyond a single news headline. Humanity created machines to make work easier. Now it faces a different question. What happens when the machine no longer makes work easier but takes it over entirely? It is an old dream that has become a new uncertainty. For generations, progress was sold as a promise that life would improve for everyone. What we are witnessing instead is a form of progress that harms first the very people whose work made it possible. The 4,800 Microsoft employees are not an abstract statistic. They are the price a company is paying to finance a technology that promises to make even more human labor unnecessary in the future.

In the end, one question remains that no balance sheet can answer. If the world's smartest companies invest hundreds of billions of dollars in a future that requires fewer and fewer people, who will ultimately be left to buy the products that future creates? Machines earn no wages and know no hunger. But they also have no purchasing power. Perhaps that is the real equation that no one is willing to confront.