This should be made clear: artificial intelligence did not create weaknesses in the exam system. She only illuminated them with the power of a thousand spotlights. It has long been pointed out that exams reward memory mastery of the material too much at the expense of the ability to think creatively, cooperate or adapt. AI has become an accelerator of this crisis, brutally invalidating the previous rules of the game.

Devaluation of restorative knowledge – The end of the era of “forge, surrender, forget”

At the heart of the current matura exam is the verification of what the student is able to reproduce. This applies both to the canon of readings in Polish, dates and facts in history, as well as formulas in science subjects. Artificial intelligence massively “commodifies” this knowledge – it reduces it to the level of a commodity available on demand, in a fraction of a second. An argumentative essay based on “Forefathers’ Eve”, an analysis of a poem, or even the solution of a complicated task in organic chemistry become tasks that the algorithm performs with increasing precision.

This phenomenon leads to a deep crisis of meaning. What is the purpose of many years of training in writing a essay according to a rigid key, if the machine does it better? What’s the point of cramming hundreds of formulas when they’re built into smart calculators? Continuing to examine reproducible knowledge in its current form becomes a kind of educational theater of the absurd – we all pretend to take part in a process whose fundamental assumptions have been undermined. The discussion between CKE and artificial intelligence must go beyond the problem of cheating and touch the very essence of what and why we check at all.

Redefining Academic Integrity: The Line Between Tool and Fraud

The entire examination system is based on the concept of “independent work”. In the age of AI, this definition becomes fluid and extremely problematic. In adult, professional life, the ability to intelligently use AI assistants to solve problems faster and more effectively is and will be a key competence. A programmer who does not use Copilot is already losing out on the market today. A marketer who ignores AI tools works slower. We, on the other hand, as part of a three-hour exam, create an artificial environment in which we prohibit the use of the most basic tools of the modern world.

This gives rise to cognitive dissonance. On the one hand, we promote digital competences. On the other hand, during the most important test in a young person’s life, we punish them for their use. This state of affairs is not sustainable in the long term. The future of the baccalaureate must take into account this new paradigm and find a way to assess how a student thinks with tools, not just what they remember without them.

How will the exams change? Evolution scenarios from gradual adaptation to radical revolution

A system with such great inertia as central exams will not change overnight. Reforming education and AI will be a process that is likely to involve several evolutionary stages.

Stage I: “AI-Resilient” exams – Increasing the importance of analytical and problem tasks

This is the easiest and most likely first step. It consists in a thorough redesign of exam sheets within the existing, analogue formula. The goal is to limit as much as possible the tasks that can be done with the “copy-paste” method with AI.

In practice, this means: Fewer commands “describe”, “list”, “characterize”. More tasks that require:

• Synthesis: Comparison of two different texts (e.g. a fragment of a novel and a scientific article) and formulation of conclusions.
• Critical Evaluation: Analyses of the source text in terms of credibility, hidden assumptions or intentions of the author.
• Creative solution to the problem: Presenting a hypothetical problem (e.g. social, ethical, design) and asking for a proposed and justifying solution.
• Personalization and Reflection: Tasks that require you to refer to your own experiences, values, and thoughts, which AI can’t genuinely do.

Stage II: “Open” exams with access to resources – Performance test

The next, logical step is to move towards exams with partial or full access to digital resources (the so-called “open-book exams”). In such a model, having knowledge becomes secondary. The key competence to be assessed is information effectiveness – the ability to instantly find the data you need, critically select it, verify it and apply it to solve a complex problem in a limited time.

• In practice, this means: The worksheets would have to consist of comprehensive case studies, which would require the student to take on the role of an analyst, designer or strategist, who must “deliver” a solution within 3 hours, using the available tools. This is a model closer to the realities of professional work, but requiring a huge change in the way tasks are formulated.

Stage III: Diversification of forms of competency validation – End of the monopoly of the written exam

The AI revolution is likely to accelerate the shift away from the monopoly of the written exam in favor of more diverse forms of skill testing.

• Increasing the role of oral exams: Direct conversation, debate, defense of a presentation – these are the forms that allow you to verify your ability to argue live, your fluency in thinking and your authentic understanding of the topic.
• Digital Portfolio as an Element of Assessment: Instead of basing the entire assessment on a single, several-hour test, universities and employers may in the future take into account a digital portfolio documenting the best projects, research work and practical achievements of a student from the entire learning period. Such a form, known from artistic studies, can be extended to other fields. Validation of competencies through documented, real achievements becomes much more credible. AI-assisted assessment can play a key role here, supporting teachers in the ongoing, objective evaluation of portfolio-based work.

How to teach when the rules of the game change?

With such deep uncertainty about the future of exams, it is natural to ask, “So how am I supposed to teach my students?” The answer is paradoxically simple and extremely liberating. The best way to prepare students for the exams of the future is… immediate cessation of teaching “for the current exams”.

The benefits of learning and learning about new AI tools in the classroom are not only higher motivation and a better understanding of the material. It is primarily a training of competencies, which will be crucial in each of the possible scenarios of exam evolution. A student who learns in lessons:

• Have an intelligent dialogue with the AI, not just give simple commands.
• Critically evaluate AI responses and verify them against external sources.
• Use AI to synthesize information and solve complex problems.
• Use technology ethically and transparently document your work…
  … is a student who will do well on the “AI-resistant” exam, on the “open” exam and will successfully build their portfolio. Artificial intelligence in schools, wisely implemented, ceases to be a problem, and becomes a solution – an incubator of skills that make the student independent of a specific test formula.

Become an architect of change, not an observer of it

The revolution in the exam system, triggered by artificial intelligence, has already begun. We can either passively wait for top-down decisions and guidelines from the CEC, or start actively shaping the future of education in our classrooms and schools today. The latter path requires courage, but above all knowledge – a deep understanding of both the opportunities and threats posed by the new technology.

Understanding these complex processes and translating them into everyday teaching practice is a huge challenge. Therefore, if you feel that you want to be not just an observer, but a conscious participant and leader of this transformation, you need solid support.

“Artificial Intelligence for Schools and Teachers” is not another textbook with a list of applications. It’s a comprehensive AI guide for education that equips you with strategic knowledge and practical tools to navigate this new reality.

Don’t let the pace of change overwhelm you. Gain knowledge that will give you confidence and allow you to turn the biggest challenge of modern education into your greatest opportunity. Prepare yourself and your students for a future that has come sooner than we expected.

Inequalities in access to technology and exam fairness

In the discussion about exams and AI , the fundamental system of social justice cannot be overlooked. The rapid development of paid, increasingly powerful language models and specialized AI tools is creating a new digital stratification. Students from wealthier homes with access to premium subscriptions (such as GPT-4 and higher) gain a significant advantage over those who can only use the free, often functionally limited versions. This advantage is manifested not only in the quality of the generated content, but also in access to advanced analytical functions, data synthesis or image interpretation. The future of the matura exam, if it is to remain an exam that equalizes the odds, must take this factor into account. This means that any attempts to incorporate technology into the exam process will have to be based on ensuring that all examinees have access to an identical, standard tool. Otherwise, the exam, instead of measuring competence, will largely measure the economic status of the parents. This issue also sheds new light on the issue of preparation for the matura exam – new, expensive courses of “prompt engineering for high school graduates” will be created, deepening the existing inequalities.