Harvard: AI Startups Hire Fewer Juniors and Bet on Experts

A working paper from Harvard Business School and INSEAD, whose results became known in July 2026, documented a structural difference between AI-native startups and traditional technology companies: they hire significantly fewer entry-level specialists, build compact flat teams, and bet almost exclusively on senior technical personnel.

Who and what researched

The authors of the report are Rembrandt Koning from Harvard Business School and Hyunjin Kim from INSEAD. They studied startups that went through the Y Combinator accelerator between 2020 and 2024, and compared them with a broader sample of technology companies from the same time period. Business Insider journalists were the first to report on the results of the working paper.

The study revealed a persistent pattern: the more densely AI is integrated into the core of a startup's business, the fewer employees it has in entry-level positions.

Key parameters of AI-native companies in the sample:

• Teams are more compact than competitors without AI at the core of the product
• Hierarchical structure is flat, with minimal number of levels
• The share of senior technical personnel is significantly above market average
• Hiring specialists without experience is a rarity, not a norm

Why AI startups do without juniors

The reason is structural, not cyclical. Tasks that were traditionally covered by entry-level specialists — basic programming, testing, data processing, preparation of technical documentation — are today largely handled by AI tools. When a model copes with routine tasks faster and cheaper, the need for employees without formed competencies drops sharply.

The hiring request itself is shifting. AI-native startups need people capable of making complex technical decisions from day one: designing system architecture, evaluating model quality, determining the boundaries of their applicability, integrating AI so that it solves real problems rather than creating an illusion of automation. Training "from scratch" in such an environment is an expensive endeavor for both the company and the employee.

The result: instead of the familiar pyramid with a wide base of juniors and a growing layer of mid-level specialists, a flat team of highly qualified specialists is formed. The company turns out to be more compact and faster, but sets a much stricter entry threshold.

What does this mean

If the pattern documented by Koning and Kim becomes the norm for the entire industry — and Y Combinator represents a fairly representative sample of promising startups — the technology sector job market is facing a serious structural shift. There will be fewer entry-level jobs for junior specialists, the qualification threshold for starting a career will rise, and the question of how to prepare young specialists to work alongside AI from day one will become more acute for universities, bootcamps, and corporate training programs.

AI changes not only the content of work — it changes the very logic of career paths.