US jobs paradox: AI-exposed roles see 46% pay surge even as AI-linked layoffs rise
The wage data do not mean AI is having little impact on employment. US employers have announced substantial numbers of layoffs explicitly linked to artificial intelligence.
- Sep 21, 2026,
- Updated Sep 21, 2026 8:53 AM IST
Artificial intelligence is beginning to reshape the US labour market in a way that is more complicated than a simple story of machines replacing workers. While companies are increasingly using AI to automate tasks and reduce headcount, workers in occupations most exposed to the technology are also seeing the fastest growth in advertised pay.
A chart shared by The Kobeissi Letter on X, based on Indeed data, shows advertised salaries in the most AI-exposed US occupations rising 46% since 2021, compared with 41% for moderately AI-exposed jobs, 39% across all posted wages and 25% for the least AI-exposed occupations.
The divergence has become particularly visible since 2024 and widened further through 2025 and 2026. The chart indexes 2021 average pay at 100; by 2026, the index had reached about 146 for highly AI-exposed jobs, compared with 141 for moderately exposed roles and 125 for the least exposed.
The Kobeissi Letter said the trend suggests employers are increasingly competing for workers with AI-related skills, adding that AI is reshaping the US compensation landscape.
AI exposure is translating into a pay premium
The underlying Indeed Hiring Lab analysis, published on September 17, found that the occupations most exposed to generative AI include software development, IT systems and support, data and analytics, marketing, and banking and finance. Less-exposed occupations include nursing, personal care, food preparation, cleaning and manufacturing.
The finding challenges one of the most common assumptions about AI: that greater exposure automatically means lower wages.
Indeed found that, after accounting for changes in the mix of occupations, AI-exposed jobs carried a 5.7% post-ChatGPT advertised-pay premium. When researchers compared individual job titles with their own historical pay, the premium remained at 4.7%. Controlling further for seniority reduced it to 2.4%, suggesting that part of the premium is connected to the growing importance of senior and specialised workers.
The premium is also not uniform. Senior roles have recorded the largest cumulative pay gap, followed by mid-level jobs, while the difference is much smaller at entry level.
That distinction matters because AI may be changing what companies want from employees rather than simply eliminating entire occupations. Employers can use AI to automate coding, research, data analysis, drafting and other repetitive tasks while simultaneously placing a higher value on workers who can supervise AI systems, integrate them into workflows and handle more complex problems.
But the other side of the AI story is layoffs
The wage data do not mean AI is having little impact on employment.
US employers have announced substantial numbers of layoffs explicitly linked to artificial intelligence. According to Challenger, Gray & Christmas, AI was cited in 116,175 announced job cuts in the US through August 2026, accounting for about 22% of all announced cuts. AI remained the leading cited reason for job cuts on a year-to-date basis, although it fell to fourth place in August, when 3,462 cuts were attributed to AI.
The number has risen sharply over the past two years. Challenger recorded 54,836 AI-related announced layoff plans in 2025. Since 2023, when the firm began separately tracking AI as a reason for job cuts, it has cited AI in more than 170,000 announced cuts.
The figures need to be read carefully. An employer citing AI does not necessarily mean AI alone caused every affected position to disappear. Companies frequently combine AI adoption with restructuring, cost reduction, changes in business strategy and broader workforce reorganisation.
Still, the pattern is visible in major corporate decisions.
In February, Block announced plans to eliminate more than 4,000 jobs, nearly half its workforce, as part of an AI-driven restructuring. CEO Jack Dorsey said AI had changed how companies could operate, allowing smaller teams to accomplish more.
Amazon also cut jobs in its artificial general intelligence group in July, following a 16,000-employee layoff announced in January. The company said the AGI cuts were part of efforts to streamline priorities and accelerate work in areas it considered most important.
Meta has also been restructuring around AI. Its Project OT initiative sought to create smaller teams supported by AI, with plans that could have reduced some teams substantially. The company ultimately scaled back parts of the plan after concerns about productivity, reliability, security and employee resistance.
Entry-level workers face different problem
The emerging labour-market divide may therefore not be simply AI jobs versus non-AI jobs. It could increasingly become a divide between workers who can use AI to amplify their productivity and those whose tasks can be automated or require fewer people.
Indeed found that software-development job postings, for example, have rebounded strongly over the past year. But 71% of the increase in software-development postings between May 2025 and May 2026 came from senior roles, while 37% came from jobs that explicitly mentioned AI in their titles.
That points to a potential pressure point for younger workers. If AI can handle a growing share of the routine tasks traditionally assigned to junior employees, companies may need fewer entry-level workers to produce the same output.
The International Labour Organization (ILO) has highlighted a similar risk. Its 2026 review of empirical evidence found that large-scale job displacement remains limited so far, but warned of growing inequality and the erosion of employment opportunities for younger workers as AI changes how work is organised.
The ILO estimates that roughly one in four workers globally are in occupations with some degree of exposure to generative AI. But it stresses that exposure should not be treated as a prediction of job losses. Most occupations contain a mixture of tasks that can and cannot be automated, meaning AI is more likely to transform many jobs than eliminate them outright.
The World Economic Forum similarly expects technological and other structural changes to create 170 million jobs globally by 2030 while displacing 92 million, producing a projected net increase of 78 million jobs. Those figures cover broader labour-market transformation, not AI alone.
For Unparalleled coverage of India's Businesses and Economy – Subscribe to Business Today Magazine
Artificial intelligence is beginning to reshape the US labour market in a way that is more complicated than a simple story of machines replacing workers. While companies are increasingly using AI to automate tasks and reduce headcount, workers in occupations most exposed to the technology are also seeing the fastest growth in advertised pay.
A chart shared by The Kobeissi Letter on X, based on Indeed data, shows advertised salaries in the most AI-exposed US occupations rising 46% since 2021, compared with 41% for moderately AI-exposed jobs, 39% across all posted wages and 25% for the least AI-exposed occupations.
The divergence has become particularly visible since 2024 and widened further through 2025 and 2026. The chart indexes 2021 average pay at 100; by 2026, the index had reached about 146 for highly AI-exposed jobs, compared with 141 for moderately exposed roles and 125 for the least exposed.
The Kobeissi Letter said the trend suggests employers are increasingly competing for workers with AI-related skills, adding that AI is reshaping the US compensation landscape.
AI exposure is translating into a pay premium
The underlying Indeed Hiring Lab analysis, published on September 17, found that the occupations most exposed to generative AI include software development, IT systems and support, data and analytics, marketing, and banking and finance. Less-exposed occupations include nursing, personal care, food preparation, cleaning and manufacturing.
The finding challenges one of the most common assumptions about AI: that greater exposure automatically means lower wages.
Indeed found that, after accounting for changes in the mix of occupations, AI-exposed jobs carried a 5.7% post-ChatGPT advertised-pay premium. When researchers compared individual job titles with their own historical pay, the premium remained at 4.7%. Controlling further for seniority reduced it to 2.4%, suggesting that part of the premium is connected to the growing importance of senior and specialised workers.
The premium is also not uniform. Senior roles have recorded the largest cumulative pay gap, followed by mid-level jobs, while the difference is much smaller at entry level.
That distinction matters because AI may be changing what companies want from employees rather than simply eliminating entire occupations. Employers can use AI to automate coding, research, data analysis, drafting and other repetitive tasks while simultaneously placing a higher value on workers who can supervise AI systems, integrate them into workflows and handle more complex problems.
But the other side of the AI story is layoffs
The wage data do not mean AI is having little impact on employment.
US employers have announced substantial numbers of layoffs explicitly linked to artificial intelligence. According to Challenger, Gray & Christmas, AI was cited in 116,175 announced job cuts in the US through August 2026, accounting for about 22% of all announced cuts. AI remained the leading cited reason for job cuts on a year-to-date basis, although it fell to fourth place in August, when 3,462 cuts were attributed to AI.
The number has risen sharply over the past two years. Challenger recorded 54,836 AI-related announced layoff plans in 2025. Since 2023, when the firm began separately tracking AI as a reason for job cuts, it has cited AI in more than 170,000 announced cuts.
The figures need to be read carefully. An employer citing AI does not necessarily mean AI alone caused every affected position to disappear. Companies frequently combine AI adoption with restructuring, cost reduction, changes in business strategy and broader workforce reorganisation.
Still, the pattern is visible in major corporate decisions.
In February, Block announced plans to eliminate more than 4,000 jobs, nearly half its workforce, as part of an AI-driven restructuring. CEO Jack Dorsey said AI had changed how companies could operate, allowing smaller teams to accomplish more.
Amazon also cut jobs in its artificial general intelligence group in July, following a 16,000-employee layoff announced in January. The company said the AGI cuts were part of efforts to streamline priorities and accelerate work in areas it considered most important.
Meta has also been restructuring around AI. Its Project OT initiative sought to create smaller teams supported by AI, with plans that could have reduced some teams substantially. The company ultimately scaled back parts of the plan after concerns about productivity, reliability, security and employee resistance.
Entry-level workers face different problem
The emerging labour-market divide may therefore not be simply AI jobs versus non-AI jobs. It could increasingly become a divide between workers who can use AI to amplify their productivity and those whose tasks can be automated or require fewer people.
Indeed found that software-development job postings, for example, have rebounded strongly over the past year. But 71% of the increase in software-development postings between May 2025 and May 2026 came from senior roles, while 37% came from jobs that explicitly mentioned AI in their titles.
That points to a potential pressure point for younger workers. If AI can handle a growing share of the routine tasks traditionally assigned to junior employees, companies may need fewer entry-level workers to produce the same output.
The International Labour Organization (ILO) has highlighted a similar risk. Its 2026 review of empirical evidence found that large-scale job displacement remains limited so far, but warned of growing inequality and the erosion of employment opportunities for younger workers as AI changes how work is organised.
The ILO estimates that roughly one in four workers globally are in occupations with some degree of exposure to generative AI. But it stresses that exposure should not be treated as a prediction of job losses. Most occupations contain a mixture of tasks that can and cannot be automated, meaning AI is more likely to transform many jobs than eliminate them outright.
The World Economic Forum similarly expects technological and other structural changes to create 170 million jobs globally by 2030 while displacing 92 million, producing a projected net increase of 78 million jobs. Those figures cover broader labour-market transformation, not AI alone.
For Unparalleled coverage of India's Businesses and Economy – Subscribe to Business Today Magazine