Accounting, Law and Business Administration: AI is not erasing these degrees, it is redefining their tasks

By Zendoric · September 8, 2026.

An article in El Liberal (Santiago del Estero, Argentina) places three university degrees at the center of the conversation about AI and employment: Accounting, Law and Business Administration. According to the text itself, an "analysis carried out by AI" —without specifying authorship, institution or methodology— ranks these degrees among those that could see some of their functions reduced by automation in the coming years. It is worth saying it just as plainly: that vagueness in the source does not invalidate the underlying thesis, but it does mean it must be read as a generic snapshot, not as a rigorous study with verifiable data.

The facts the article does provide are concrete and plausible. In Accounting, it notes that current tools already produce balance sheets, calculate taxes, organize budgets and process large volumes of information, routine tasks that used to require hours of human work. In Law, it points to case-law research, contract review and document analysis as the most automatable fronts, while interpretation, negotiation and strategy in complex cases still require a professional. In Business Administration, the focus is on data analysis, trend detection, inventory management and decision-making support.

This connects directly with something we have already documented with more solid sources here: analyzing the impact of AI on employment sector by sector, we saw that in Law the firm's pyramid narrows at the base —junior research work is the first to go— while the value of expert judgment and of the client relationship rises. In Business Administration the trend is identical: fewer purely administrative profiles, more analysts and managers who know how to operate these tools, with increasingly flat organizations. El Liberal's article reaches the same conclusion by a less rigorous route, but the destination is the same.

Our reading is that the relevant question for anyone choosing a degree today is not "will this profession disappear?" but "what part of this profession will still need a human a few years from now?". That is the real shift: from the routine, codifiable task —the one a language model already performs reasonably well today— toward judgment, negotiation, the relationship of trust with the client and ultimate responsibility for a decision. It is exactly the pattern that repeats in every sector we have analyzed: automation does not empty out professions, it reorders them from the inside.

In the short term this is partly bad news, and it should be said without sugarcoating: the most routine entry-level positions —the junior accountant doing repetitive bookkeeping entries, the paralegal who only searches case law— are precisely the ones that feel the pressure first, and they are the ones that traditionally served as the training ground for tomorrow's senior professional. Universities and employers have yet to work out how to build expert judgment if you remove the rungs where that craft used to be learned through repetition. It is a real transition problem, not a minor detail.

In the long term, however, the reading we hold at Zendoric is that freeing accountants, lawyers and managers from the most mechanical part of their work opens up space for them to devote themselves to what truly adds human value: fine-grained advice, strategy, the relationship of trust. It is the same logic we apply to the AI-driven economy as a whole: the more abundance and efficiency these tools generate in repetitive tasks, the more room is left for people to work on what really matters to them. The key, for anyone deciding what to study today, is not to avoid these degrees but to enter them knowing that the value no longer lies in the mechanical task, but in everything the machine still cannot do.