AI Upends the Elite Career Formula as Law Degrees and MBAs Lose Their Status as ‘Guaranteed Tickets’ to High Incomes

Automation Spreads From Case-Law Research to Financial Analysis
Junior Roles Shrink as Apprenticeship-Based Career Ladders Collapse
Degrees Lose Their Luster as Empathy and Verification Skills Set New Market Value

Law schools and Master of Business Administration (MBA) programs have long been regarded as a “guaranteed ticket” to high-paying professions. A widely held belief suggested that enduring steep tuition costs and years of education would secure stable careers as lawyers or consultants, along with high salaries and social status. That formula for success is now faltering as artificial intelligence (AI) rapidly absorbs work traditionally assigned to entry-level employees, including case-law research, contract review, market research, and financial analysis. The premium once conferred by professional licenses and academic degrees is diminishing, while the ability to detect AI errors and interpret the complex dynamics of people and markets is emerging as the new determinant of professional value.

Law School Popularity Plummets

According to the Korean Association of Law Schools on Aug. 7, the number of applicants for the Legal Education Eligibility Test (LEET), the first gateway to law school admission, fell from 19,400 for the 2025 academic year to 19,057 for 2026. The figure plunged 9.8% year over year to 17,184 for the 2027 academic year. Only 15,560 applicants actually appeared for the examination last month. With law schools admitting 2,000 students annually, the applicant pool remains 8.6 times larger than total enrollment capacity, indicating that competition remains intense. Yet the nine-year growth streak has clearly ended. Waning confidence in the scarcity value and income premium historically conferred by a law license appears to be prompting prospective students to scrutinize the cost-effectiveness of law school from the outset.

Law school requires three years of substantial tuition and living expenses. Opportunity costs rise further when forgone earnings during enrollment are included. Graduates must pass the bar examination to qualify as lawyers, but the pass rate has recently hovered around 50%. Even after qualification, income disparities widen sharply depending on whether a lawyer secures a position at a major law firm or corporate legal department. The supply of lawyers has also expanded rapidly. South Korea produces approximately 1,700 new lawyers each year, while the legal market has failed to grow sufficiently to absorb them.

AI Takes Over Entry-Level Legal Work

The rise of AI has added another layer of uncertainty. Case-law research, evidence classification, contract review, and initial drafting—work once assigned to junior lawyers—are already being automated. “Most major law firms use proprietary AI programs for tasks such as case-law research and evidence organization,” said a partner at a major South Korean law firm. “The former apprenticeship model, in which junior lawyers analyzed hundreds of pages of legal briefs and received feedback, has become unnecessary.”

A Thomson Reuters (TRI) survey of 1,514 professionals across 27 countries found that the share of law firms using generative AI rose from 28% in 2025 to 41% this year. Adoption among corporate legal departments more than doubled from 23% to 47%. Among legal professionals, AI was used for document review by 77%, legal research by 74%, document summarization by 74%, and drafting briefs and memoranda by 59%. Tasks that once required several junior lawyers to work long hours while developing practical expertise can now be completed with AI in a fraction of the time.

“AI Outperforms Law Professors’ Answers”

The quality of AI-generated answers is also improving rapidly. AI performance in the delivery of legal knowledge has already reached a level capable of competing with law school faculty. In a working paper released in May by Stanford Law School researchers, 16 US law professors created 40 contract-law questions, wrote their own answers, and compared them with AI-generated responses. Across 2,918 blind evaluations, the professors selected the AI answers 75.3% of the time. AI responses were deemed potentially harmful to learning in 3.5% of cases, compared with 12.1% for professors’ answers.

The study, however, was confined to short-answer questions covered in contract-law courses. Courtroom advocacy, fact-finding, client counseling, and ethical responsibility were excluded from the experiment. The scope is too narrow to support the conclusion that AI can perform every function of a lawyer, yet the technology has clearly secured considerable competitiveness in locating and organizing case law and statutory provisions.

Knowledge for AI; Judgment, Reasoning, and Accountability for Humans

The spread of AI is reshaping the role of legal professionals. Empathy carries greater value, particularly in work involving direct interaction with people. Clients facing divorce, inheritance disputes, criminal proceedings, or corporate conflicts rarely seek legal answers alone. Lawyers must present practical options that account for emotions, financial circumstances, family relationships, and competing interests within companies.

Another determinant of a lawyer’s market value will be the capacity to reason through the unstated implications embedded in legal provisions. The law cannot anticipate every real-world dispute, and a single factual distinction can produce a different outcome under the same precedent. Lawyers must establish priorities among competing principles and weave fragmented facts into a coherent argument capable of persuading the court.

Humans also retain responsibility for AI-generated results. AI can cite nonexistent precedents or produce persuasive answers while omitting material facts. This phenomenon is known as “hallucination.” AI errors are often presented in polished language, making them difficult to detect without professional expertise. The foremost human responsibility is to identify errors and falsehoods in AI-generated output and repeatedly make fine-grained corrections to improve its accuracy.

An understanding of AI capabilities is equally essential. Professionals must accurately determine what AI can and cannot do. The ability to collect, organize, and identify patterns in data to construct a broader picture also remains important. AI produces high-quality output only from high-quality input. Bringing fragmented information together meaningfully, supplying context, and designing the overall analytical framework remain human responsibilities.

The End of Apprenticeship Training

US law schools are already changing their teaching methods in response. Stanford Law School requires students to grade AI-generated legal answers and identify fabricated citations and omitted arguments. The University of Chicago Law School has identified advocacy, strategic judgment, critical thinking, and client relationship-building as essential capabilities in the AI era.

These changes are also disrupting law firms’ longstanding workforce models. Major firms have traditionally maintained pyramid-shaped organizations in which large cohorts of junior lawyers devoted extensive hours to research and document review, with a select few eventually promoted to partner. Clients were billed for the hours logged by junior lawyers. That model has become increasingly difficult to sustain as AI completes in minutes work that previously took hours.

In its March analysis of “Legalweek 2026,” the Thomson Reuters Institute concluded that the foundation of the law-firm pyramid was weakening as AI reduced foundational work for junior lawyers and companies handled more legal work internally. Pressure on hourly billing has intensified as clients resist paying under the old model for work whose required time has been sharply reduced by AI. Discussions are already spreading across the US legal market over flat-fee and value-based billing models that price services according to case complexity, outcomes, and value delivered to clients instead of hours worked.

Table 1. Changes in the US MBA and Consulting Recruitment Market Amid AI Expansion

MBA Programs Also Confront Declining Applications

MBA programs, which alongside law schools form one of the two main pillars of the professional graduate education market, face the same disruption. According to The Wall Street Journal (WSJ), major US business schools confronting declining applications are fighting to attract students with extraordinary incentives, including tuition discounts of up to 50%. Behind these steep discounts lies a severe decline in demand for traditional two-year, full-time MBA programs. MBA demand has typically moved inversely to the labor market, declining during hiring booms and increasing during downturns.

The current employment environment follows a different pattern. As anxiety grows that the rapid adoption and development of AI could threaten their jobs, workers are increasingly choosing to hold tightly to their current positions instead of resigning to change careers or pursue further education—a phenomenon known as “job hugging.” Companies have historically hired large numbers of MBA graduates, assigned them to data collection, financial analysis, and strategic report writing, and developed them into managers through internal promotion contests. Generative AI can now produce industry research, company valuation analysis, and presentation drafts at low cost, sharply eroding the cost-effectiveness of this development model.

Employment prospects have also dimmed. A recent Graduate Management Admission Council (GMAC) survey of more than 600 corporate recruiters found that one in three companies had begun restructuring recruitment plans because of AI. Respondents said some entry-level positions were already being replaced by the technology. “The lower rungs of the ladder on which young graduates relied are disappearing,” said Gad Levanon, chief economist at the labor-focused think tank Burning Glass Institute.

Compensation structures have also changed. According to the Financial Times (FT), McKinsey and Boston Consulting Group (BCG), among other leading consulting firms, have frozen starting compensation for university and MBA graduates for a third consecutive year. Compensation remained at $135,000–$140,000 for undergraduate degree holders and $270,000–$285,000 for MBA graduates. As AI raises productivity in research and report writing traditionally performed by junior consultants, companies are reducing large-scale entry-level recruitment and placing greater emphasis on securing experienced professionals and industry specialists. The consulting industry’s pyramid-shaped workforce model, in which large numbers of junior employees support a small group of partners, is also being destabilized.

The Limits of MBA Knowledge as Doctoral-Style Thinking Sets Market Value

In the AI era, the ability to rapidly organize existing knowledge is becoming less scarce. Generative AI can summarize market data, compare competitors, and produce initial drafts of financial models and business strategies. At the same time, AI job-market data indicate that demand is shifting unevenly toward professionals capable of working with models and technical systems, even as many routine white-collar tasks lose value. This development is placing pressure on MBA education, which emphasizes the broad acquisition of management knowledge applicable across multiple industries. Meanwhile, the ability to define problems, formulate hypotheses, and test them against evidence is gaining value.

The recent focus on “doctoral-style thinking” reflects this shift. Doctoral training emphasizes identifying weaknesses in existing research, examining a single field over an extended period, and repeatedly testing the validity of conclusions. This research-oriented mindset is necessary to identify errors, biases, and leaps in causality within AI-generated answers. The growing market for specialized AI and data strategy MBA programs reflects the same pressure, although the technical depth of these programs varies considerably.

One institutional response can be seen at the Swiss Institute of Artificial Intelligence’s Gordon School of Business and Artificial Intelligence (GSB). Rather than treating AI as a limited elective layer added to a conventional management curriculum, its STEM AI MBA tracks in AI/Big Data and AI/Finance are structured around mathematical, statistical, and data-science foundations linked to business and institutional decision-making. The approach reflects a broader change in the labor market: managers increasingly need sufficient technical depth to test model assumptions, challenge automated conclusions, and communicate with specialists—not merely consume AI-generated summaries.

The international scientific journal Nature has also called for doctoral education to be redesigned for an environment in which AI performs data analysis and writing tasks. Columbia Business School likewise concluded that MBA education in the AI era must shift from describing phenomena and proposing solutions toward deeper training in identifying root causes and exercising judgment.