Poland's Law School Deans Warn AI Is Undermining Legal Education

More than 90 deans and representatives of law faculties from across Poland met from September 2 to 4, 2026, in Olsztyn to discuss how artificial intelligence is changing the training of future lawyers. Dr. hab. Krystyna Szczechowicz, dean of the Faculty of Law and Administration at the University of Warmia and Mazury, warned that the problem is no longer theoretical - practicing lawyers are already submitting court filings full of rulings invented by AI.

An Unprecedented Challenge

The conference, held under the banner "Challenges of Contemporary Science in an Era of Technological and Social Change," brought together deans from nearly all of Poland's law faculties. Alongside discussions on the funding of law studies and the system for evaluating academic research, a separate session was devoted to how artificial intelligence is affecting the process of educating and assessing students' knowledge.

As people who train young lawyers in an era of technological, social, and economic change, we face enormous challenges. Challenges like these have never existed before. - Dr. hab. Krystyna Szczechowicz, dean of the Faculty of Law and Administration at UWM in Olsztyn

Szczechowicz stressed that universities find themselves in a split situation. On one hand, artificial intelligence is becoming a permanent part of legal practice and cannot be ignored; on the other, law faculties must protect the quality of academic research and the copyright of work produced at universities.

On one hand, we have a situation where artificial intelligence is being introduced into our reality, and on the other, we must protect the quality of scientific research and intellectual property rights. On yet another hand, we're aware that we can't do without using artificial intelligence, but it can't replace independent thinking. - Dr. hab. Krystyna Szczechowicz, dean of the Faculty of Law and Administration at UWM in Olsztyn

Fabricated Rulings in Court Files

The most concrete part of Dean Szczechowicz's remarks concerned examples from professional practice, not just student coursework. As she explained, there have already been cases in which practicing lawyers used artificial intelligence uncritically and submitted court filings citing Supreme Court rulings that never existed, because the AI model had simply invented them.

Szczechowicz noted that this type of error is not accidental but stems from the very design of large language models, which, when uncertain of an answer, generate plausible-sounding but false content to satisfy the user's expectations. The consequences of such mistakes are no longer borne only by students but also by practicing lawyers - Polish professional bodies are conducting disciplinary proceedings against some of them, which in extreme cases can end in the loss of the right to practice law.

Teaching Critical Thinking

According to the UWM dean, the answer to these risks cannot be a ban on using AI tools, since that would be fighting windmills given how widespread they already are in legal work. Instead, law faculties need to teach students how to verify what a model produces before it ends up in any filing or academic paper.

We must teach verification, critical thinking, today. We have to teach young people to use AI under control. - Dr. hab. Krystyna Szczechowicz, dean of the Faculty of Law and Administration at UWM in Olsztyn

In practice, this means, among other things, analyzing the same legal cases in two parallel tracks - once using traditional methods, once with AI tools - and then jointly identifying the moments where the model "distorted or stated something untrue." Szczechowicz allows limited use of AI by students, for example to check the linguistic correctness of their papers, but stresses that submitted work must be the result of independent thinking, not content generated by a model.

Consequences for Degrees and Titles

The dean also pointed out that obtaining an academic degree or title based on work actually written by artificial intelligence, rather than independently, constitutes unlawful acquisition of a title and can carry criminal consequences. That is why a separate panel at the congress was devoted to the possible uses of AI in academic publishing, peer review, and the preparation of theses and dissertations - universities want to work out common standards before the problem grows to the scale already seen in professional practice.

For the Polish legal market, this is a signal that the problem of AI-fabricated citations and rulings, until now known mainly from isolated disciplinary and court cases, has already made it onto the academic community's agenda as a systemic challenge. Law faculties, which train future judges, attorneys, and legal counsels, acknowledge that the question is no longer whether students will use AI, but how to teach them to do so safely before they enter the profession.

The congress in Olsztyn was held under the patronage of the Polish Press Agency (PAP) and funded by the Ministry of Science and Higher Education, showing that the topic of AI in legal education has already reached the level of government interest in higher education. Future meetings of deans are expected to serve as an opportunity to assess whether the rules on AI use in law studies now being developed actually reduce the number of errors that later show up in professional practice.