How Artificial Intelligence Might Eliminate 4 Million Jobs Annually—Without Collapsing the Job Market

Main Insights

• According to Goldman Sachs economists, artificial intelligence could slow job growth by eliminating between 1 million and 4 million positions annually, but is also expected to generate even more new roles.
• Occupations previously considered at risk from automation, such as fitness trainers and real estate professionals, have actually seen continued expansion.
• AI may open up new employment opportunities for those who use these technologies, and could increase demand for service industry jobs as incomes rise.

While AI has the potential to replace certain jobs, it is also likely to create a greater number of new positions in the workforce.

This optimistic outlook comes from Goldman Sachs analysts, who recently released a report suggesting that artificial intelligence will not trigger widespread joblessness. Instead, they predict that although millions of jobs may be automated, a similar or greater number of roles will emerge, helping to maintain low unemployment rates.

The report serves as a counterpoint to Citrini Research, which warned that AI could displace so many workers that it would destabilize the economy. Concerns about job losses due to AI contributed to recent fluctuations in the stock market, according to experts.

Economic Implications

If the projections from Goldman Sachs hold true, the widespread anxiety about AI causing massive unemployment may be exaggerated.

Goldman Sachs acknowledges that AI will be highly disruptive, with estimates suggesting that between 1 million and just over 4 million jobs could be lost each year in the coming period. Nevertheless, Joseph Briggs, a global economist at Goldman, maintains that these changes are unlikely to cause a significant rise in unemployment.

Briggs notes, "The U.S. economy generates over 30 million new jobs annually, and technological innovation has historically driven long-term employment growth. We anticipate that AI will follow this trend, creating new opportunities even as it replaces some existing roles. As a result, we do not foresee a widespread loss of jobs."

He further explains that AI can stimulate job creation by enhancing productivity, enabling new roles that involve working with AI, and increasing the need for service workers as overall incomes grow.

Briggs' research into employment data reveals little evidence that AI has led to significant job losses outside of a few areas, such as software engineering. He also points out that, historically, even when technology can perform certain tasks, human workers often remain essential in those same fields.

For example, employment among fitness instructors—once thought to be at risk due to workout videos and, more recently, fitness apps—and real estate agents, whose roles have been impacted by online platforms, has actually outpaced overall job growth both in recent years and over the past quarter-century.

Looking Ahead

Goldman Sachs envisions a future for AI that mirrors previous technological revolutions. Citing MIT professor David Autor's research, they highlight that 60% of the jobs available in 2018 did not exist in 1940, illustrating how innovation continually reshapes the labor market.