AI employment impact has begun to show cracks, Goldman Sachs: 15 million American workers may experience job transitions

```

A few weeks ago, Goldman Sachs CEO David Solomon was downplaying concerns about mass unemployment caused by AI, but the bank's research department soon released its most systematic report to date on AI’s impact on jobs, with a warning conclusion: The first measurable cracks in the employment market caused by AI have already appeared.

The report shows that employment in industries with high AI penetration in the US is already contracting, and layoffs explicitly attributed to AI are accelerating, while the cost of programming-type AI agents is now significantly lower than that of human developers. Goldman Sachs estimates that around 15 million American workers may ultimately face job transfers during the entire AI technology transition period, though the bank expects that many of these positions will be filled over time by new professions.

The timing of this report is quite delicate—it comes from the same institution whose CEO just publicly claimed that the employment impact of AI is overhyped. The report does not proclaim that the "AI apocalypse" has arrived, but it clearly states that structural pressure on the job market is moving from predictions to reality, especially in industries where AI applications are already at scale, rather than an overall spread across the economy.

Employment is Already Contracting in Industries with High AI Penetration

Goldman Sachs data shows that in industries where AI use cases are mature and implemented, the number of employees is decreasing at a net rate of about 11,000 people per month. The proportion of tech industry jobs remains below its long-term trend line, and the wage growth rates in multiple non-tech and tech service industries have also fallen below the baseline trend from 2015 to 2019.

The report cites US Bureau of Labor Statistics data, noting that departures from job growth trends are especially concentrated in industries with established AI use cases, rather than being a universal phenomenon across the economy. Goldman Sachs emphasizes that "skill enhancement is just the first step in driving changes in the labor market," and that it takes a process for technical capabilities to translate into real employment impact—but that process has already begun.

The Proportion of Layoffs Attributed to AI is Rising Rapidly

Data from Challenger, Gray & Christmas shows that in recent months, high-profile tech companies have explicitly attributed layoffs to AI with increasing frequency, and the number of such layoff announcements continues to rise.

Goldman Sachs’ quantitative analysis further reveals the dual impact of productivity gains driven by technology: for every 1% increase in technology-driven productivity, the unemployment rate rises by 0.3 to 0.4 percentage points; meanwhile, over a two-year cumulative period, the job attrition rate increases by 0.5 to 0.6 percentage points. The report also notes that workers replaced by technology face sustained pressure on their incomes, and post-reemployment salary levels are unlikely to recover to their original level.

The Cost of AI Programming Agents is Now Lower Than That of Human Developers

When comparing the costs of AI and humans, the Goldman Sachs report points out that AI programming agents now have an economic advantage over human labor, while other types of jobs are still subject to more constraints.

Goldman Sachs believes that the key driver behind the expansion of economically rational automation is not only the continuous improvement in model capability but also the ongoing decline in token costs. As inference costs are further reduced, the scope of economically viable automated jobs will steadily expand, meaning that the types of jobs currently being impacted may only be the beginning.

Whether New Jobs Can Outpace the Rate of Job Loss is the Core Issue

The Goldman Sachs report also acknowledges historical trends: technological progress will create new jobs in the long run. The bank expects that among the 15 million workers replaced, many will eventually find opportunities in emerging professions.

However, the report also cites economist Daron Acemoglu's warning: "Economics does not have any general law that guarantees the rate of job creation will necessarily match the rate of job destruction." Notably, the unemployment rate for current college graduates is already higher than other educational groups—a phenomenon regarded as an early signal that the impact of AI is spreading into entry-level white-collar jobs. For investors, the real question is whether the labor market can create new jobs faster than AI eliminates old ones.

Risk Warning and DisclaimerThe market has risks and investment should be made with caution. This article does not constitute personal investment advice, nor does it take into account any user's specific investment objectives, financial situation, or needs. Users should consider whether any opinions, views, or conclusions in this article are suitable for their specific circumstances. All investments based on this are at the user's own risk. ```