Is the first rung of the legal career ladder disappearing?

“If AI primarily focuses on entry-level work, the bottom rungs of the career ladder may crumble.”

More experienced employees seem to be able to breathe a sigh of relief. For those just starting out, the labor market looks significantly less promising. New research from Harvard shows that generative AI is a form of seniority-biased technological change: a technological development that primarily replaces junior employees, while seniors remain relatively unscathed.

The study, based on millions of CVs and job postings from 285,000 American companies, shows that since early 2023, the number of junior hires at AI-adopting companies has dropped sharply. Senior staff, on the other hand, remains in high demand. This brings a fundamental question to the fore: if AI takes over entry-level roles, how will the professionals of tomorrow be trained?

What exactly did Harvard investigate?

The study combines data from LinkedIn and Revelio Labs and analyzed:

• 62 million CVs and 245 million job vacancies,
• from 285,000 American companies,
• over a ten-year period (2015–2025),
• distinguishing between junior and senior positions,
• and a clever way to measure actual AI adoption, based on job postings for ‘AI integrator’ roles.

This allowed researchers to measure the effect of actual AI implementation within organizations for the first time, rather than just looking at sectors with theoretically high exposure to AI.

The blow falls on entry-level starters

The results leave little room for doubt. Since early 2023, the number of junior jobs at AI-adopting companies has decreased by 7.7%. Not because people were laid off, but because fewer junior positions are being opened. Instead of recruiting juniors, companies are filling these tasks with AI.

At the same time, the number of senior positions continues to grow, as does the number of promotions from junior to senior within companies actively using AI. The message is clear: those already on the inside get opportunities. But those yet to start are increasingly facing closed doors.

Not just tech; retail and services are also affected

Although you might expect AI effects first in the technology sector, the impact is broader. In sectors such as retail, customer contact, and professional services, the number of junior positions is rapidly declining.

The numbers do not lie:

• In the wholesale and retail sector, the number of junior hires is falling by nearly 40%.
• In other sectors, such as professional services, information, and education, the decline ranges between 15% and 30%, depending on the sector.
• In contrast, demand for senior staff remains stable and even rises by 5% to 10% in some sectors compared to non-adopting companies.

Routine tasks once performed by starters, such as communication, documentation, or reporting, are increasingly being automated with generative AI tools.

The U-curve: who you are (or where you studied) suddenly matters a great deal

A striking aspect of the research is the analysis of educational background.

The researchers categorize universities into five quality levels. What are the findings?

• Alumni from top institutions (Tier 1) remain relatively well protected.
• Students from the cheapest and lowest-rated institutions (Tier 5) retain their positions due to their low costs.
• But the middle group, graduates from solid, well-regarded universities, is losing the most work.

This U-shaped distribution shows how AI not only affects jobs but also puts social mobility under pressure.

What does this mean for legal careers?

Although the Harvard study does not focus specifically on legal professionals, the parallels with legal practice are unmistakable.

After all, the classic model of legal training revolves around learning in practice:

• Junior lawyers start by analyzing files, drafting memos, and conducting legal research.
• It is precisely these types of cognitively challenging yet routine tasks that are susceptible to automation by AI.

If these tasks disappear, an essential learning path also vanishes. In the long run, this could lead to a shortage of well-trained senior lawyers with sufficient practical experience. For firms that rely on the ‘pyramid structure’ with high intake at the bottom, this is a significant risk.

AI is changing not just work, but careers as well

The researchers warn that the disappearance of entry-level roles is more than a temporary problem. Those who break off the bottom rungs of the career ladder risk having no one left at the top in the long run.

Conclusion: being prepared for a quieter intake

The Harvard study makes it clear that AI is changing not only the workplace but also the structure of careers and talent development.

For organizations, this means a strategic reconsideration:

• How do you develop the professionals of tomorrow if you no longer hire starters?
• How do you prevent a situation where, in five years, you have no seniors with sufficient experience?
• And what does this require from educational programs, HR policies, and leadership?

For lawyers, advisors, and policymakers, there is a clear mandate: keep the influx of talent in focus.

Because without a solid base, there is no top—but perhaps that is exactly the crux of the matter. If generative AI undermines the classic pyramid structure of many firms, a fundamental question arises: will we soon be building a new model? One with fewer layers, leaner teams, and different forms of learning and career progression?

Anyone who takes this signal seriously will have to think about alternative models for talent development, because if the entry point disappears, AI forces us not only to work differently but also to train differently.