Norwegian secondary teachers’ perceptions of artificial intelligence in summative assessments

Traditional summative assessment practices in Norwegian secondary schools have faced criticism for potential biases, inconsistencies in grading reliability, and heavy workload demands on teachers, prompting interest in whether artificial intelligence (AI) can enhance fairness, objectivity, and efficiency. This study examined secondary school teachers’ perceptions of AI as a tool for summative assessments and identified factors influencing their acceptance of its integration. A cross-sectional design targeted teachers from five socioeconomically and culturally diverse secondary schools in Oslo. Convenience sampling yielded 223 valid responses from paper-based questionnaires (response rate ~87%). Relationships among latent variables were tested using structural equation modelling: belief in AI for assessment enhancement, digital instructional efficacy, workload/time pressure, and preference for traditional teaching methods. Key findings revealed that digital instructional efficacy was positively associated with greater openness to AI-driven assessments (β = .38), supporting its role in promoting perceived efficiency and objectivity. Preference for traditional methods showed a moderate negative correlation with AI acceptance (β = −.18), while workload pressures had a negligible direct impact on AI beliefs (β = −.02) but indirectly reinforced reliance on conventional approaches. The results indicate a tension between progressive technological adoption and established pedagogical ideologies, suggesting that AI integration requires strategies that build teachers’ digital confidence while respecting traditional practices and managing workload to avoid entrenchment in familiar methods. Limitations include the Oslo-centric sample and reliance on self-reported data; future research should incorporate longitudinal designs, qualitative methods, and broader regional representation.