Goldman Sachs argues fears of an AI-driven "job apocalypse" are overstated. While AI will significantly reshape the labor market, the firm expects disruption to unfold gradually rather than trigger mass unemployment.
Key takeaways:
- Around 15 million U.S. workers (9% of the workforce) could be displaced over the next decade in Goldman's base case, but the process is expected to be gradual.
- AI could automate roughly 25% of U.S. work hours, particularly in white-collar jobs such as finance, law, consulting, customer service, software development, and graphic design.
- The labor market impact is already visible but limited. AI-linked job losses are concentrated in tech, creative, and knowledge-based industries, while overall U.S. unemployment has not risen materially because of AI.
- Productivity gains are expected to outweigh job losses over time. Goldman argues AI should boost economic growth, lower costs, and create entirely new occupations, similar to previous technological revolutions.
- The biggest risk is the speed of adoption. If AI replaces workers much faster than expected, unemployment could rise more sharply before new jobs are created.
- Younger and knowledge-sector workers are likely to experience the greatest disruption initially, while demand for construction, power infrastructure, and data-center jobs is expected to increase alongside AI investment.
Bottom line
Goldman Sachs' conclusion is that AI is likely to transform the labor market rather than destroy it. The transition could be painful for millions of workers, but the firm expects productivity gains and the creation of new industries to offset much of the long-term employment impact, provided the adjustment happens over several years rather than all at once.
