AI Job Growth includes ChatGPT-Fueled Surge Amid Overall Employment Slowdowns

DeepSeek to Accelerate Trend Across Industries, According to UMD-LinkUp AI Maps

COLLEGE PARK, Md., Feb. 3, 2025 /PRNewswire/ -- Since ChatGPT's late-2022 launch, AI job postings in the United States "have ramped up dramatically, from 29,509 in Q4 2022 to 49,577 in Q4 2024, an uptick of 68 percent," according to new analysis at UMD-LinkUp AI Maps, the world's first tool for mapping the creation of jobs requiring artificial intelligence skills and co-produced by researchers at the University of Maryland's Robert H. School of Business.

The findings are the subject of the white paper "Diffusion of AI Jobs Across Economic Sectors."

In the same period, all U.S. job postings overall declined by 17 percent. "While this decline does not suggest that companies are employing fewer people, it does suggest that companies are hiring new staff at a slower pace," write the researchers, co-led by Smith's Anil K. Gupta, Michael Dingman Chair and professor of Strategy, Globalization and Entrepreneurship.

"Our data are consistent with a broader softening in the labor market," they add, citing a recent Minneapolis Fed report that "the unemployment rate is unequivocally on the rise, from a historic low of 3.4 percent in 2023 to 4.3 percent in July 2024." It is also notable that "the decline in IT job postings is much starker"—from 354,070 in Q4 2022 to 258,706 in Q4 2024, a decrease of 27%.

The researchers identify AI jobs as those requiring AI skills, while IT jobs comprise a broad group of computer- and math-related occupations.

Given the divergent upward trend for AI job postings, "there is clear evidence of a strong ChatGPT effect," the researchers write.

And moving forward, "the DeepSeek phenomenon will accelerate these trends," adds Gupta. "Irrespective of what happens with DeepSeek, its emergence has vindicated with full force Mark Zuckerberg's bet on open-source LLMs."

He explains: "As open-source models keep gaining share, the market for foundation models will become commoditized and their costs and prices will decline. This will lead to an even faster deployment of AI technology in every industry, be it software development, technical services, banking, insurance, manufacturing or agriculture."

While it's a bit early for this effect to show up in job postings data, Gupta adds, "we anticipate seeing these numbers over the next few months."

Sector-Level Analysis

On the key measure of AI Jobs Intensity (i.e., share of postings for AI jobs versus for all jobs), three sectors stand out: "information" (at 3.24 percent), "professional, scientific, and technical services" (at 2.40 percent), and "finance and insurance" (at 1.54 percent). As a benchmark, the AI Jobs Intensity for the U.S. economy as a whole is 0.72 percent. According to Gupta, "the reason is very clear, in these three sectors, all work is knowledge work, and thus amenable to augmentation or substitution by AI."

The contrast with two sectors at the other extreme is stark. Together, "health care and social assistance" and "accommodation and food services" account for 28 percent of all job postings, but only 1.5 percent of AI job postings. The AI Jobs Intensity of these two sectors is 0.05 percent or lower. Why? Because these two sectors require highly complex physical work, not yet amenable to AI or robotics.

Professional, Scientific, and Technical Services. The latest findings show "dramatic growth" in the AI job-postings share of the "professional, scientific, and technical services" sector in the overall U.S. economy – from 13.5 percent in 2018-Q1 to 24.0 percent in 2024-Q4. Gupta explains: "While practically all sectors of the economy have embraced AI, it appears that many companies in the historically less digitized sectors have chosen to outsource their AI expertise to consulting firms. These include the likes of Accenture, Deloitte, EY, PwC, Booz Allen, and others which belong to this sector."

Information. "Software publishers" dominate this sector, accounting for over 43 percent of all job postings in the sector. This subsector's share of AI job postings is even larger – 62 percent.

Manufacturing. Within the vast and diverse "manufacturing" sector, the "computer and electronic products" subsector vastly outshines the other manufacturing subsectors. This subsector accounts for 46 percent of AI job postings in the sector, even though its share of all job postings is only 15 percent.

Finance and Insurance. Since the launch of ChatGPT in 2022-Q4, AI postings increased sharply by 62 percent versus a very sharp 33 percent decline in IT postings. As in the other sectors, here as well, there is strong evidence that companies are investing in AI jobs at the expense of more general IT jobs. In terms of AI Jobs Intensity, the three main subsectors – "insurance carriers," "commercial banking," and "investment banking" – quite similar.

Retail. Unlike the "information" sector, "retail" employs vast numbers for mostly physical work in brick-and-mortar outlets and in fulfillment centers for purely online retailers, like Amazon's retail business. Thus, the AI Jobs Intensity of this sector trails that for the US economy. However, this sector's AI-to-IT jobs intensity is significantly higher than for the U.S. economy. On this measure, Walmart's numbers are close to those for Amazon retail (18.45% versus 22.66%) and way ahead of all other retailers (at 8.98%). "These data reflect Walmart's very aggressive investments in AI, including at its technology hubs in Silicon Valley and India," Gupta says.

UMD-LinkUp AI Maps is published in partnership with job-data firm LinkUp (a subsidiary of GlobalData PLC) and consulting firm Outrigger Group. Smith School researchers joining Gupta in the project are Dean's Professor of Information Systems Siva Viswanathan, Associate Professor of Information Systems Kunpeng Zhang and doctoral student Hanwen Shi.

Contact: Greg Muraski at [email protected]

SOURCE University of Maryland's Robert H. Smith School of Business