Microsoft (MSFT.US) offers voluntary buyout packages to about 60,000 employees: Structural divergence in U.S. employment under AI transformation

According to Woofun AI, the U.S. labor market in the fall of 2026 is undergoing a profound structural divergence: beneath the surface of initial jobless claims remaining at a near 57-year low of 194,000 and an overall unemployment rate of just 4.2%, only 29,000 non-farm jobs were added in September, with both hiring and quit rates simultaneously falling into stagnation.

This phenomenon of "low layoffs and weak hiring" is not a simple cyclical fluctuation, but a deep transformation triggered by AI-driven efficiency restructuring, marked by Microsoft (MSFT.US) offering voluntary buyout packages to about 60,000 employees and American Express (AXP.US) announcing a reduction of roughly 5% (about 3,500 people) of its global workforce.

A sharp contrast has emerged between the calm of macro data and the radical nature of corporate micro-behaviors. The number of continued jobless claims is around 1.7 million, indicating that large-scale mandatory layoffs are still rare, and companies prefer to adjust structures through natural attrition, such as encouraging early retirement, accepting voluntary departures, and not filling the vacancies. Microsoft's voluntary buyout plan covers about a quarter of its global workforce as eligible, essentially constituting voluntary exits; while American Express directly emphasized preparing for AI deployment to enhance organizational agility and decision-making efficiency.

Data compiled by Woofun AI shows that this model of advancing AI efficiency restructuring through voluntary buyouts and attrition is reshaping the labor cost structure of tech giants, making current employees relatively secure, while job seekers face an unusually challenging market environment.

The segmentation of the job market further intensifies the reversal between industries. The reemployment cycle has lengthened, with the 20-24 age group, fresh graduates, and inexperienced entry-level job seekers being hit the hardest, their unemployment rates significantly higher than the overall level. The reason is that companies are conservative in their hiring intentions, prioritizing retention of experienced employees, and using AI tools to let a few highly skilled workers take on tasks previously done by multiple junior employees, thereby shrinking entry-level positions.

Meanwhile, the historical pattern where rate-sensitive sectors like construction and manufacturing led recessions through early contraction has seen a partial reversal. Non-residential construction employment has expanded significantly due to data center construction, boosting durable consumer goods manufacturing (such as cooling equipment and power equipment);

Despite moderate contraction in residential construction, overall employment in construction and manufacturing continues to post net growth, becoming a key factor supporting recent economic resilience, while service and white-collar jobs are facing different pressure dynamics.

Productivity gains have not translated into broad-based real income growth, but have instead exacerbated distribution imbalances. Information technology, finance, and professional services are the industries with the highest AI adoption rates. Census Bureau surveys show that real output per worker in these sectors saw a clear upward inflection from late 2022 to early 2023. These high adoption-rate industries together account for about 15% of the workforce. Preliminary estimates show that 80% to 85% of productivity gains in AI-related industries currently flow into corporate profits, with only a small amount converted into real wages. As a result, nominal wages have risen by about 3.0% annually, while the consumer price index is around 3.7%, leading to a slight decline in real wages. Workers long accustomed to real wage growth of about 2% to 2.5% now feel greater pressure from major expenses such as housing and automobiles, with the Michigan Consumer Sentiment Index remaining at low levels for an extended period.

In this self-reinforcing cycle, companies reinvest extra profits into AI infrastructure, while small businesses (which account for over 40% of employment and GDP) face greater uncertainty. The NFIB survey shows that in September 2026, its optimism index is slightly below the long-term average, the uncertainty index is notably above historical norms, and factors such as labor quality, inflation, tariff policy volatility, and energy price shocks have made hiring plans more cautious.

Looking ahead, Federal Reserve policy remains a key variable. Current leading indicators, aside from residential construction, are generally improving, and corporate profitability remains resilient, providing the economy with some capacity to absorb rate hikes. However, if policy tightening continues and transmits to significant weakening of employment in construction and manufacturing, cyclical pressures could compound existing structural frictions. The continued moderate decline in residential construction employment warrants close attention, as it has historically been a leading indicator of recession. Real wage growth is unlikely to improve markedly in 2027, and deep-seated structural factors, including negative externalities caused by excessive government debt and spending, are restraining real income growth in the private sector. The market is currently in transition: blue-collar jobs are supported by data center construction, while entry-level white-collar positions are under pressure. Policy priorities should include lowering the threshold for retraining and improving mechanisms for creating entry-level jobs to avoid long-term suppression of private sector real incomes by expanding the scale of government. The definition of "security" in the labor market is being redefined, shifting from simply whether one is unemployed to whether one can continue to gain accumulative experience and real income growth in a rapidly changing technological environment.