ACADEMIC INTEGRITY AS A COMPONENT OF BIOETHICS: NEW REQUIREMENTS FOR TEACHERS AND STUDENTS OF HIGHER EDUCATION IN THE ERA OF ARTIFICIAL INTELLIGENCE

The issue of utilizing Artificial Intelligence (AI) in medicine has become widespread in scholarly literature from 2023 to 2025.Given that large generative models like ChatGPT can successfully pass licensed medical exams such as the USMLE, intense discussions have emerged regarding the blurring lines between human knowledge and algorithmic capabilities, and the adherence to bioethical principles and academic integrity.The objective of this work was to conduct a bioethical analysis of the risks associated with the implementation of generative AI in the research and medical activities of $\text{PhD}$ students and their supervisors, and to substantiate the need to transform the concept of academic integrity from the principle of "text originality" to the principle of "responsibility for data verification."This shift is seen as a necessary condition for maintaining the fundamental bioethical rule of "do no harm" (nonmaleficence) in clinical practice.Currently, evidence-based medicine is founded on the axiom that published data are based on real clinical observations and experiments.However, large language models (LLMs) operate not with facts as such, but with probabilistic connections between words.The existing problems pose a significant danger: the fabrication of bibliographic references and scientific falsification undermine trust in the scientific communication system, distort meta-analytic data and systematic reviews, and contribute to the proliferation of works containing fundamental data interpretation errors.Such dishonesty poses a direct threat to patient safety.Журнал «Перспективи та інновації науки» № 12(58) 2025 1883Another critical aspect is the crisis of the concept of "authorship."Authorship is not merely a technical function but an ethical one, implying the ability to vouch for data reliability, defend research results in scientific discussion, and bear responsibility in case falsifications are uncovered.Bioethics establishes a clear boundary: AI may assist in writing, but responsibility for what is written lies exclusively with the human author.Furthermore, there is the risk of raising a generation of physicians suffering from "automation bias": they may become proficient in using tools to find answers but gradually lose the capacity for independent clinical analysis in situations where the algorithm is unavailable or erroneous.Conclusions.The rapid implementation of generative AI into the scientific and medical domain has caused an irreversible transformation of the very notion of academic integrity.It has evolved from a purely administrative category into a fully-fledged bioethical category, directly linked to the principle of non-maleficence.LLMs are prone to technical "hallucinations" and the replication of algorithmic bias, therefore, a researcher's integrity today is defined not by the refusal to use AI, but by the ability to take responsibility for the rigorous verification of every generated statement.AI cannot become a guarantor of patient safety; it will always remain merely a tool, not a co-author of the scientist and physician.This necessitates a fundamental change in the educational paradigm: shifting the focus from combating plagiarism to fostering the cognitive autonomy of the future specialist.The prospects for the further development of bioethics in this matter lie in finding a balance where PhD student and instructor utilize the power of AI for data analysis while preserving critical thinking and empathy as the exclusive prerogatives of the human being.Only such a balance can prevent professional degradation, avoid "deskilling," and ensure that medicine does not transform into an automated technical process without a "human face."