Goldman Sachs Says AI Will Displace 15 Million US Workers
"If we look back over the last 80 years, around 85% of job growth has been driven by the technological creation of new positions."
Joseph Briggs, who leads Goldman Sachs Research's global economics team, said on the bank's Exchanges podcast that AI will displace approximately 9% of the US workforce. That is almost 15 million workers.
The displacement is already visible in the data. In sectors where AI tools are actively in use, including tech, management consulting, and graphic design, Briggs estimates AI is reducing monthly employment growth by 10,000 to 15,000 jobs.
Briggs' argument about what it means is more nuanced. He pushed back directly against what he described as a fixation among tech leaders on jobs destroyed while ignoring jobs created.
"If we look back over the last 80 years, around 85% of job growth has been driven by the technological creation of new positions," he said. The US labor market creates and destroys roughly 30 million jobs annually, a churn rate large enough, Briggs argued, that even a modest acceleration in job creation would be sufficient to reabsorb the workers AI displaces.
The Counterargument From MIT
Neil Thompson of MIT, also a guest on the Exchanges podcast, offered a slower timeline but not a fundamentally different conclusion. His argument was that AI's technical capabilities and its actual economic deployment are two separate questions.
Deployment requires the right data, cost efficiency at scale, and the absence of regulatory barriers. Those conditions will not arrive simultaneously across the economy.
Thompson also challenged the framing of total job displacement. Most jobs, he argued, will be partly automated rather than fully replaced, and the outcome depends on which specific tasks machines take over.
He used GPS as his analogy: when GPS automated taxi drivers' expert knowledge of city routes, wages fell but the number of drivers grew significantly. AI, in his framing, is a "rising tide" that workers can see coming and adapt to, not a "crashing wave" that sweeps them away without warning.
The June 2026 US jobs report gave both arguments immediate context. The economy added just 57,000 jobs, roughly half what economists expected, with April and May revised down by a combined 74,000.
The unemployment rate dipped to 4.2%, but largely because workers exited the labor force rather than because hiring increased. Whether those numbers represent the early signs of AI-driven displacement, a temporary economic slowdown, or both remains an open question.
Key Takeaways
- Goldman Sachs predicts AI will displace approximately 15 million US workers, or 9% of the workforce.
- AI is already impacting job growth, reducing employment growth by 10,000 to 15,000 jobs monthly in certain sectors.
- Historical data shows 85% of job growth stems from technological advancements creating new positions.
- Experts emphasize the need to balance concerns over job loss with potential for new job creation.
- AI deployment faces challenges, requiring data, cost efficiency, and regulatory clarity to fully integrate.