Will Generative AI make good jobs harder to find?

Youth unemployment and underemployment are growing challenges worldwide, especially for college graduates. Families have long invested in higher education for their children, dreaming of careers in fields that require specific skills and expertise like law, banking, engineering, or diplomacy. College enrollment rates surged, tripling from 14% in 1990 to 42% in 2022. Yet, many of these ambitions have not materialized. In 2023, one in five young people globally were not in employment, education, or training, with women making up two-thirds of this group. In the United States, over half of recent college graduates are in jobs that do not require a college degree.

The lack of productive and stable white-collar jobs for those with university diplomas is especially acute in developing economies, where the creation of such jobs lags. In low- and lower-middle-income countries, more than a fifth of those under age 30 who have postsecondary qualifications are unemployed, much higher than those with a basic education. In Sub-Saharan Africa, nearly three out of four young adult workers aged 25 to 29 were in insecure jobs – meaning they were either self-employed or in temporary positions. In the Arab States and North Africa, one in three economically active youth are unemployed. China also experienced a significant spike in its official youth unemployment rate in recent years, surpassing 20% in June 2023.    

As this crisis unfolds, a new player has entered the job market: Generative Artificial Intelligence (GenAI). Will it be the solution we’ve been waiting for, or will it exacerbate an already dire situation? Our recent working paper examines how GenAI might affect the economy – its growth, changes in industry structure, and international production patterns. Here’s what we discovered:

1. GenAI Predominantly Boosts Productivity in High-Skill Services

GenAI is set to have a significant impact on white-collar jobs within high-skill services sectors, which are typically filled by those with a college education. Unlike previous waves of digital technologies, which primarily expedited routine tasks or made predictions by recognizing data patterns, GenAI’s ability to synthesize and generate ideas and content intersects with a substantial portion of tasks in white-collar occupations. Several studies consistently indicate that the jobs most exposed to GenAI are concentrated in high-skill services (Eloundou et al. 2023; Gmyrek, Berg, and Bescond 2023; World Economic Forum 2023; Melina et al. 2024). Our findings reveal that finance and insurance, ICT services, and professional services—three high-skill, high-income, and highly digitalized industries—are the most vulnerable to the influence of GenAI.