AI Kills Jobs, Stanford Study Finds, Especially For Young People

The Impact of AI on Employment: A Growing Concern

A recent study by Stanford researchers has shed light on the significant impact of Artificial Intelligence (AI) on employment in the United States. The study, which analyzed data from millions of workers across thousands of companies, found that AI is causing job loss, particularly among entry-level workers. The hardest-hit areas are occupations where AI automates tasks, rather than augmenting them, with early-career workers experiencing a 13% relative decline in employment.

The study used data from ADP, the largest U.S. payroll processor, to analyze employment patterns from January 2021 through July 2025. During this period, overall employment grew, but employment growth among young workers began to stagnate starting in late 2022, around the time ChatGPT was released to the public. This coincides with a growing swell in AI innovation that has continued to accelerate since then.

The Disproportionate Impact on Young Workers

Younger workers are disproportionately impacted by AI-driven job loss, with entry-level jobs in software development and customer service being particularly affected. According to Ford CEO Jim Farley, half of all white-collar jobs are at risk. This is a huge risk, as entry-level jobs have always been the proving ground for future leaders, and if AI erases that foundation, companies risk losing an entire generation of managers and project leaders.

The reality is that there are now numerous AI software development tools, such as GitHub Copilot, Cursor, Replit, CodeGPT, Bolt, and Amazon Q Developer, which can automate tasks and potentially replace human workers. Additionally, a search for “AI customer service” on popular business software comparison site G2 returns over 1,800 results, highlighting the widespread adoption of AI in this sector.

The Key Factor: Automation vs. Augmentation

The key factor in whether AI kills jobs is how it is applied. Employment declines are concentrated in occupations where AI automates tasks, rather than augmenting them. In contrast, in occupations or roles where AI augments tasks, entry-level employment remained stable or even grew. The rate of improvement in AI tools is staggering, with AI systems able to solve just 4.4% of coding problems in a widely used benchmark in 2023, but hitting 71.7% in 2024.

However, where AI is being used to augment work instead of replace workers, it’s not having a downward impact on wages, the study says. Some jobs are also more resistant to AI-driven job loss, particularly those that require interpersonal interaction, creative skills, or physical activities. Lower AI-exposure jobs, such as caring for others, are less likely to be automated and may even see growth in employment.

A Shift in Perspective: Collaborating with AI

The key to thriving in this transition may lie in how we view jobs and AI. According to Sri Ambati, CEO of AI agent company H2O, jobs are disappearing where AI automates work, but growing where AI augments human capabilities. Early-career workers need to shift from competing against AI to collaborating with it more effectively. This may involve developing skills that complement AI, such as communication, creativity, and problem-solving.

Interestingly, English and liberal arts majors are able to communicate better with AI systems than more technical people, and can sometimes get better results thanks to more detailed and context-rich prompts. As Ambati says, “The workers thriving in this transition aren’t trying to outrun automation; they’re learning to run with it.” By adopting this mindset, we can harness the potential of AI to augment human capabilities, rather than replacing them.