Ethics of AI Use in Scientific Knowledge Development: A Philosophy of Science Perspective on Students at SMAN 1 Ukui

Although artificial intelligence (AI) has received considerable attention in education, research on ethical AI use in scientific knowledge development among secondary school students remains limited, particularly from a philosophy of science perspective. This study examined students’ use of AI to develop scientific knowledge, their perceptions of information validity, and ethical considerations concerning academic integrity through epistemological and axiological perspectives. A descriptive quantitative approach supplemented by qualitative data was employed. Surveys involved 120 Grade XI and XII students at SMAN 1 Ukui, while semi-structured interviews involved 12 purposively selected student and teacher participants. Data were analyzed using descriptive statistics and thematic analysis. AI was frequently used for idea exploration (74.2%), literature synthesis (68.3%), and report writing (62.5%), whereas only 22.5% of students reported regular or frequent use for fact verification and source checking. Furthermore, 58.3% associated convincing AI-generated language with factual correctness, and 51.7% perceived AI as more objective than human reasoning in scientific data analysis. Although 81.7% considered the fabrication of research data using AI unethical, views on undisclosed AI-generated writing and the delegation of scientific conclusions to AI remained divided. These findings highlight the importance of critical information verification, independent scientific reasoning, transparency, and academic responsibility in AI-assisted knowledge development. By integrating epistemological and axiological perspectives, the study contributes to understanding AI ethics in secondary education and supports the development of school-level guidelines and AI literacy education that promote academic integrity. Research across multiple schools could extend these insights to different educational contexts.