Jobs with the greatest potential for AI-assisted productivity gains also face the highest projected job losses, according to a new index from Digital Planet at the Fletcher School at Tufts University.
THE US Risk Index for AI Jobs ranks 784 occupations in the United States, 530 metropolitan areas, 50 states, and 20 industry sectors based on their vulnerability to AI-related job losses.
All figures are model projections based on AI adoption scenarios, not actual layoffs or job changes. The middle scenario estimates that 9.3 million jobs are at risk, ranging from 2.7 million to 19.5 million depending on the speed of AI adoption.
Which jobs face the highest projected risk
Writers and authors top the list of at-risk professions 57%. Computer programmers and web and digital interface designers follow 55% each. The editors are at 54%and web developers from 46%.
Market research analysts and marketers face a 35% job loss rate. Public relations specialists are at 37%. News analysts, reporters and journalists face 35% risk.
Previous analyses, like the Anthropogenic economic index and that of Stanford “Canaries in the coal mine,» measured the extent to which jobs are accessible AI. This analysis goes further by estimating the likelihood that this exposure will result in a projected job loss.
Increase and risk of loss go hand in hand
The authors call the link between jobs that benefit from AI-generated productivity gains and those that risk losing their jobs the “augment-displacement link.”
When AI increases the efficiency of individual workers, businesses can produce the same result with fewer employees. This primarily affects entry-level and lower-seniority positions, as companies may reduce hiring rather than layoffs.
Writing, programming, web design, technical writing, and data analysis are the fields where this trend is most evident. Tasks in these domains are cognitive, language-intensive, and sufficiently structured to be handled by large language models.
By industry
The average vulnerability across all sectors is approximately 6%. The sectors with the highest projected job losses are information (18%), Finance and Insurance (16%), and professional, scientific and technical services (16%).
Software developers, management analysts, and market research analysts face the largest total revenue losses. These three roles combine high salaries and large workforces, accounting for a significant portion of the projected $757 billion in total annual revenue at risk.
What the analysis does not include
Note that job creation effects are not included in this version. The authors intend to add this data in future updates as they gather more evidence.
Additionally, regulatory constraints, union bargaining power, and professional licensing requirements that could help slow job losses in some sectors are not part of this analysis. The authors emphasize that their predictions are based on different scenarios rather than being definitive.
Why it matters
Digital professionals assume that using AI to increase productivity protects their jobs. However, this data challenges this idea.
SAY had already covered this tension in 2023 when Dr. Craig Froehle of the University of Cincinnati warned that companies that did not invest in reskilling their employees would see their turnover costs double. The Tufts data quantifies the specific professions on which this pressure is increasing.
Looking to the future
Updates to the US AI Jobs Risk Index will be made as AI capabilities and labor market conditions evolve. The authors mention that future versions will attempt to include data on job creation as well as loss estimates, providing a more comprehensive view of the overall impact of AI on employment.
THE methodology is available on the Digital Planet website, which also links to a data download page.
Featured Image: Rudall30/Shutterstock
