Abstract:

This study examines how teachers and future teachers in compulsory education perceive and use generative artificial intelligence (AI) in their professional practice, with the aim of supporting the design of a targeted AI tool for education. In February and March 2026, we conducted an online survey among 159 respondents, including both teachers and students in teacher education programs. The questionnaire focused on the current use of AI tools and large language models (LLMs), self-assessment of competencies, perceived barriers, and areas where artificial intelligence could support educational activities.

The findings indicate a high degree of adoption of AI and LLMs, as well as predominantly positive attitudes toward these technologies. More than half of the respondents report using AI tools regularly (weekly or more often), and approximately one-quarter use them occasionally. Nearly 60% rate their knowledge as good or very good, while only a small proportion report minimal or no competence. Attitudes are generally favorable: more than four-fifths of respondents express a very positive or somewhat positive opinion on the use of AI in teaching, while openly negative attitudes are negligible. The data also suggest generational differences: younger respondents more frequently report regular use and higher levels of self-confidence, while older age groups express greater caution and perceive more barriers.

Among the most frequently mentioned tools are conversational chatbots (e.g., ChatGPT-type systems), AI-powered platforms for creating visual materials, translators, and language proofreading tools. Respondents see the greatest potential for AI in supporting lesson preparation (creating worksheets and other materials), generating ideas and inspiration for teaching, language editing and translation of materials, and reducing administrative burdens. In contrast, more sensitive areas, such as assessing students and providing them with feedback or communicating with parents, are viewed with greater caution.

The results show that there is significant demand in school practice for an AI tool that would be systematically adapted to the conditions of the Czech education system. The proposed assistant should primarily relieve teachers of routine tasks related to lesson preparation and administration, thereby allowing them to focus more on their actual teaching work and interaction with students. The generational differences found, the identified types of use, and the specific areas where teachers perceive the greatest benefits of AI can be directly used as a basis for specifying the functionalities and user interface of the future tool. The study thus provides a practically oriented framework for the design, pilot testing, and potential international expansion of the proposed system aimed at supporting teachers.

Keywords:

AI, Large Language Models, education, AI tool.