Opportunities and Challenges of AI Language Models in Higher Education for Sustainable Development

Abstract

This article examines the contemporary opportunities and limitations of using large language models
(LLMs), including ChatGPT, in higher education and scientific research. It outlines the technological
foundations of LLMs, highlighting their capabilities for context-aware dialogue, language synthesis,
automated assessment, and personalization of learning pathways, including applications in language
learning and intercultural communication that can support learner autonomy and communicative
competence. The study emphasizes the potential of AI to enhance the quality of education, support
pedagogical decision-making, and improve the management of educational processes. At the same
time, key risks are identified, including informational biases, reliance on training data, the potential
generation of inaccurate content, threats to privacy, and challenges to academic integrity. Ethical
considerations are discussed, focusing on algorithmic transparency, data security, researcher
accountability, and the prevention of discriminatory effects. The article also presents key strategies for
addressing these challenges, including the development of information and ethical literacy, the
establishment of transparent university policies, clarification of scientific publication requirements,
and implementation of guidelines for responsible LLM use. The study concludes that effective
integration of LLMs into academic environments requires a balanced combination of innovative
potential and ethical safeguards to ensure the integrity of education and scientific research.

Keywords: large language models (LLMs), higher education, personalized learning, academic integrity, artificial intelligence, digital literacy, sustainable development, ethical AI, communication