AI in student work: not only allowed, but encouraged

In response to the development of artificial intelligence and the new challenges of modern education, ELTE's Faculty of Economics decided in September 2024 to launch a special preparatory course aimed at enabling students to learn to use AI consciously and responsibly. The first step was training the faculty's lecturers, which is now being followed by the introduction of the potential applications of AI to more than 7,000 students. The course, which will be integrated into the Project Assignment and Thesis course starting in Spring 2025, will introduce students to the workings, uses and limits of generative AI, in particular large language models (LLMs), while encouraging them to use the new technology with critical thinking and in a legally compliant manner. The digital course material – available in both English and Hungarian – will be accessible to all students of the Faculty, and the use of AI within the University and Faculty's regulatory framework will be open to those at an advanced stage of thesis preparation.

Practical focus and the role of prompting techniques

One of the key elements of the nearly four-hour introductory training is demonstrating the possibilities of using generative AI in student work, in particular in the preparation and development of project assignments and theses. The aim of AI applications should by no means be to enable students to produce their work in a fully automated way, but to support them in the brainstorming, research and editing processes.

For example, AI can help with literature review, editing and language optimisation tasks, or even with condensing and rewording a section. Students will learn about the potential, but also the limitations and risks of the most widely used AI applications (ChatGPT, Gemini, Claude, Perplexity, Consensus, DeepL, Google Translate, eTranslate) for text production and translation in higher education.

Another key element of the training will be to learn the technique of prompting, and to learn how to communicate with generative AI in word processing. Students will learn how to give precise and structured instructions to AI so that it generates the most relevant and informative results possible. The course will introduce prompting methods such as STAR (Situation, Task, Appearance, Refine) and CISCO (Context, Intent, Style, Commands, Outcome), which allow users to control the structure, depth and style of the responses they receive in order to achieve the most optimal results.

Particular attention will be paid to iterative prompting, which aims to teach students to continuously refine and check AI-generated responses to ensure accurate and contextually relevant results. In using AI, students should learn that the system's answers are not always perfect and should be evaluated with a critical eye and clarified with additional questions if necessary.

The training will illustrate the potential applications through many examples and encourage students to try all the possibilities, and gain as much personal experience as possible regarding how each system works.

Regulated and ethical use of AI

ELTE GTK aims to demonstrate AI’s potential and effectiveness and focus on its ethical and legal aspects. In line with the ELTE Student Requirements Framework (Requirements) and the ELTE Code of Ethics, the Faculty establishes clear rules to ensure that AI does not compromise the credibility of academic work.

One of the most important principles is that the unmodified reproduction of AI-generated texts is prohibited, as the content produced in this way cannot be considered as the student's own intellectual product. Accordingly, AI should not be used to automatically generate complete papers, but as a complementary tool to support individual work, for example in brainstorming, preparing research, refining the language of texts or structuring arguments. The undisclosed use of the AI, the complete adoption of a text generated by AI or the use of AI-generated content that exceeds the rules of citation is considered plagiarism according to the Requirements which may entail consequences: in the most serious cases, the thesis may be classified as invalid and the award of a mark of merit must be refused [Requirements 74/C (1)].

The regulations specifically state that students must clearly indicate if they have used AI in certain sections of their work. The University specifies the manner and form in which this must be documented. If a student uses AI in their work, they must indicate this in a statement which is a mandatory part of the project assignment and thesis templates, and must also include a summary table of the purpose and extent of use, linked to each phase of their academic work.

The use of AI can not replace the critical analysis and verification of sources, which is a fundamental requirement of academic work, and the content generated by AI should always be checked against reliable sources. Validation is essential, as one of the major challenges of generative AI is that the model sometimes produces inaccurate information - a phenomenon known as 'hallucination'. AI is not a traditional search engine, but a statistical language model that generates its answers based on probability, so it is particularly important that students always verify the information they receive with reliable sources.

The ethics of using AI is not only an individual responsibility of the students, but also a key concern for the university. The role of tutors and mentors in this process is crucial, as they help students to understand how AI can be used responsibly, how the content it generates can be verified and how to ensure that the work produced remains academically acceptable. ELTE GTK's goal is that the integration of AI in education is not just a technological innovation, but a conscious and controlled process that supports student development without compromising academic standards and requirements.

Developing a new mindset – and looking ahead

As AI-based tools become more prevalent in the world of business and work, employers are beginning to expect AI literacy, and those who have this knowledge and can use it responsibly and effectively will have a significant advantage in their first job.

However, using AI requires not only technological skills but also a new, critical and interactive mindset. ELTE GTK’s goal is for students to engage with AI not as passive consumers, but as active and responsible users. They should not only understand how to use it for their academic work, but also acquire the skills for confident workplace use.

Illustration: Midjourney