AI startups hire 15% fewer juniors and 20% more seniors, Harvard study shows

AI-Native Startups Reshape Hiring Patterns

AI startups are fundamentally changing how tech companies build their teams, according to a Harvard study that examined Y Combinator startups from 2020 to 2024 alongside a broader set of U.S. venture-backed firms 1

1

. Researchers Rembrand Koning and Hyunjin Kim from Harvard Business School and INSEAD found that AI-native startups are approximately 25% smaller than their non-AI peers while achieving comparable valuations, implying significantly more value created per employee 1

1

3

3

.

The Harvard Business School and INSEAD study defines AI-native startups by two critical shifts: using AI internally to boost employee productivity in tasks like coding, sales, and design, and embedding AI directly into products so customers can automate work that previously required human teams 3

3

. This dual approach to internal productivity and product development is reshaping workforce composition across the startup ecosystem.

Fewer Junior Workers, More Senior Talent Dominate Teams

The numbers paint a stark picture of changing hiring trends in the AI job market. AI-native startups employ roughly 15% fewer entry-level employees and 15% fewer managers compared to traditional startups 1

1

2

2

. Meanwhile, the share of more senior talent runs approximately 20% higher at these companies 2

2

. These firms also employ 13% more engineers than non-AI startups, reflecting the technical expertise required to build and deploy AI systems 1

1

2

2

.

AI startups have created flatter organizational structures with fewer management layers, placing senior employees in technical positions that require deep expertise rather than supervisory roles 2

2

. The study also found that AI companies raise 20% more capital per employee compared to non-AI firms and tend to have higher valuations per employee 2

2

.

Labor Market Disparities Deepen for Entry-Level Employees

The workers AI-native startups do hire skew heavily toward a particular profile. "These workers are especially likely to be graduates from elite institutions, concentrated in Silicon Valley, and male," the authors wrote 1

1

. This pattern cuts against the optimistic narrative that AI would democratize opportunity by allowing junior talent to compete with more experienced workers through AI-assisted productivity tools.

The broader tech sector mirrors these trends. Gen Z workers ages 21 to 25 comprised 15% of the workforce at large public tech companies in January 2023, but by August 2025, that percentage dropped to just 6.8% 2

2

. At private companies, young workers declined from 9.3% to 6.8% during the same period 2

2

. The average age of employees at large public tech companies rose from 34.3 years in 2023 to 39.4 years in 2025 2

2

.

AI Adoption Risks Widening Performance Gaps

The researchers warn that if AI accelerates learning for those who already use it effectively, "differential adoption rates may translate into widening performance gaps" 1

1

. This concern applies both to workers within firms and to the entrepreneurs who found them. As Matt Schulman, founder and CEO of Pave, explained to Fortune: "If you're 35 or 40 years old, you're pretty established in your career; you have skills that you know cannot yet be disrupted by AI. There's still a lot of human judgment when you're operating at the more senior level" 2

2

.

The findings echo what's already visible across the tech sector, where major companies like Meta and Microsoft have cut 23,000 roles as AI spending hits records 1

1

. Meanwhile, demand for senior AI expertise remains intense, with AWS investing $1 billion into forward-deployed AI engineers 1

1

. Recent graduates now make up just 7% of new hires at major tech companies 1

1

.

If AI-native startups continue attracting investment at current rates, the authors suggest that opportunity concentration could accelerate rather than reverse. The technology may flatten hierarchies inside companies while simultaneously steepening the climb to get into them, creating a paradox where AI improves productivity for those already inside elite organizations while raising barriers for those trying to enter the field.