Artificial intelligence is starting to put pressure on some parts of the labor market, with entry-level workers appearing particularly exposed, according to new research from Goldman Sachs.
Goldman Sachs Group (NYSE: GS) released the research Wednesday, finding that industries with greater exposure to AI have generally seen weaker job openings growth since 2022. Employment in call centers, software publishing, management consulting and advertising services has also fallen below historical trends, while AI adoption across major developed economies has reached roughly 15% to 20%, reported CNBC.
AI Pressure Is Showing Up in Some Industries
The effects are not spread evenly across the labor market. Goldman found that information and communication services, one of the industries most exposed to AI, have experienced slower employment growth across several developed economies since 2022.
Call centers showed some of the clearest weakness. According to the research, employment in the industry is 39% below trend in the U.S., 33% below trend in Canada and 27% below trend in Germany.
Goldman said the pattern indicates that AI-related employment pressures are already visible in industries where tools capable of automating work are available.
Entry-Level Workers Face a Bigger Challenge
Goldman analyzed employment growth across more than 800 occupations and found that AI-related headwinds were strongest among entry-level workers.
The finding lines up with recent research from Harvard Business School and INSEAD. A study of AI-native startups found that these companies were about 25% smaller than comparable startups, employed roughly 15% fewer entry-level workers and managers, and had a share of senior workers about 20% higher.
The researchers said AI-native companies were using the technology both to make employees more productive and to build AI directly into their products. That allows some companies to operate with smaller teams while maintaining similar valuations to traditional startups.
The trend could make it harder for younger workers to enter some fields if companies increasingly use AI to handle tasks traditionally assigned to junior employees.
AI Has Not Triggered a Broad Job Collapse
The Goldman findings do not necessarily mean the economy is heading toward widespread AI-driven unemployment.
A recent Bank of America analysis found little evidence that AI has so far caused a broad employment collapse across the U.S. economy. Industries with the highest AI exposure have seen employment largely move sideways since ChatGPT launched, while less-exposed industries grew about 2%.
Bank of America economist Stephen Juneau said, "AI replaces tasks not occupations," suggesting that workers may use AI to complete parts of their jobs faster without eliminating the entire position.
At the same time, some white-collar and entry-level roles are showing signs of pressure, while construction, manufacturing and other industries tied to AI infrastructure are creating new demand for workers.
Companies Are Still Hiring Alongside AI
Recent hiring trends also show that companies are not uniformly replacing workers with AI.
The shift is also reaching junior roles. Alphabet Inc. (NASDAQ: GOOG) (NASDAQ: GOOGL) expects to continue hiring in AI and cloud computing, while CSX Corp. (NASDAQ: CSX) expects its train and engine workforce to increase modestly. Booz Allen Hamilton Holding Corp. (NYSE: BAH) also plans to accelerate hiring after cutting thousands of jobs last year.
The developments suggest companies may be using AI to change how workers perform their jobs rather than simply removing positions. That could create a labor market where fewer junior employees are needed for certain tasks while demand rises for workers with specialized skills.
Why AI's Cost Still Matters
Economist Steve Hanke has offered a more skeptical view of AI-driven job displacement.
In an August interview, Hanke argued that replacing workers with AI on a massive scale remains difficult because the technology requires significant amounts of electricity, water, computing power and physical infrastructure. "Businesses will not be firing everybody and replacing them with AI," he said.
Hanke's argument provides a counterpoint to the latest Goldman findings. While Goldman sees measurable labor-market pressure in certain industries and among entry-level workers, Hanke argues that the cost of deploying AI at scale could limit how quickly businesses replace human workers.