New research from Goldman Sachs and Morgan Stanley is putting hard numbers on what many workers have felt anecdotally: artificial intelligence is already reshaping the American labor market, and the pain is not evenly distributed.
A Goldman Sachs U.S. Daily report authored by economist Elsie Peng finds that AI is eliminating roughly 25,000 jobs per month in roles where the technology can substitute for human labor. At the same time, AI is creating approximately 9,000 positions in occupations where it augments rather than replaces workers — yielding a net loss of about 16,000 jobs every month over the past year.
Who Is Getting Hit
The displacement is concentrated in routine, automatable positions: insurance claims clerks, bill collectors, data entry operators, customer service representatives, legal support staff, and billing roles. These are precisely the entry points where younger workers traditionally begin their careers.
Goldman's analysis shows the unemployment rate gap between workers under 30 and experienced employees aged 31–50 has widened significantly in AI-exposed occupations. The wage picture is equally stark — a one standard-deviation increase in AI substitution exposure widens the entry-level-to-experienced wage gap by approximately 3.3 percentage points.
The Double-Edged Sword
Goldman Sachs' own framework quantifies this duality: AI substitution has added roughly 0.16 percentage points to unemployment, while augmentation has offset that by 0.06 percentage points, yielding a net increase of about 0.10 percentage points to the overall unemployment rate.
Morgan Stanley's parallel analysis reaches a similar topline conclusion but frames it more cautiously. The firm agrees that AI's net labor market impact has been modestly negative so far, with sharper dislocations in specific sectors.
"Gen AI's impact on labor demand is double-edged: The same technology that can automate tasks can also augment labor, raising productivity," Morgan Stanley's research notes.
Methodology and Framework
Goldman's team combined AI exposure scores with a complementarity index developed by IMF economists to separate substitution risk from augmentation potential. This distinction matters: roles where AI handles most core tasks face elimination, while roles where AI enhances productivity alongside irreplaceable human skills are growing.
What This Means Going Forward
The data arrives at a politically charged moment. OpenAI released its own 13-page policy paper days earlier proposing public wealth funds, subsidized four-day workweeks, and tax reforms to cushion AI's economic blow — proposals that critics argue amount to the industry writing its own regulatory playbook.
For employers and policymakers, the Goldman and Morgan Stanley reports offer a clear takeaway: the disruption is no longer hypothetical. Entry-level pipelines in white-collar industries need rethinking, and workforce retraining programs must catch up to a technology that is already moving faster than the institutions designed to manage it.
