AI-native startups are hiring fewer entry-level workers, Harvard study finds

A recent study from Harvard Business School and INSEAD has uncovered a striking trend in AI-native startups: these companies are opting for smaller, more specialized teams and hiring fewer entry-level employees compared to their non-AI counterparts. The research, titled "AI-Native Firms," analyzed startups from Y Combinator between 2020 and 2024, as well as a broader array of U.S. venture-backed firms that secured initial funding during the same timeframe. These AI-native startups are defined by two key shifts in productivity. Firstly, they leverage AI internally to enhance employee efficiency—assisting in tasks like coding, sales, design, and project management. Secondly, they embed AI directly into their products, enabling customers to automate tasks that would have previously required human intervention. The findings indicate that AI-native startups are approximately 25% smaller in size, with a 13% higher proportion of engineers. Notably, the representation of entry-level workers and managers is around 15% lower than in non-AI-native startups. This challenges the widely held belief that the AI revolution is democratizing job opportunities, as the study suggests that entry-level employees are taking on greater responsibilities sooner, aided by AI technologies. Additionally, the paper highlights an increasing demand for seasoned professionals in these firms. The proportion of senior workers at AI-native startups is 20% greater, attracting a specific demographic of highly skilled individuals, often graduates from prestigious institutions and predominantly male. The authors express concern that this trend may exacerbate existing demographic disparities in the workforce. "AI-driven companies are employing smaller teams made up of exceptionally qualified and technical workers. This may lead to a concentration of opportunities among those already well-equipped with credentials," the researchers noted. They warn that if the use of AI tools accelerates learning for certain individuals, it could widen performance gaps both within companies and among entrepreneurs in the industry.