INSTITUTIONAL COMMUNICATION, POLICY FAMILIARITY, AND PERCEIVED AI-POLICY QUALITY IN UKRAINIAN HIGHER EDUCATION: A CROSS-SECTIONAL SURVEY

Generative artificial intelligence has entered higher-education practice faster than many institutions have developed stable governance and implementation support. This study examines GenAI use and the relationship between policy communication, policy familiarity, and educators’ perceptions of institutional AI-policy quality in Ukrainian higher education. A cross-sectional survey of 902 educators was analysed using descriptive, psychometric, non-parametric, ordinal, and multivariable methods. AI use was nearly universal, while institutional communication remained uneven: faculty or departmental communication and training were the most common information sources, and 18.2% of respondents were unaware of any policy. The five-item perceived policy-quality composite demonstrated very high internal consistency, although confirmatory model fit was mixed. In adjusted analyses, greater policy familiarity, regular institutional updates, participation in training or workshops, and official institutional email communication were associated with higher perceived policy quality. Informal discussion, independent searching, and faculty-level communication showed no independent association after adjustment. Because specific institutions were not identified, the findings represent respondent-level associations rather than university-level effects. The results suggest that formal AI-policy adoption should be supported by sustained communication, practical training, and clear operational guidance.