"When Algorithms Teach": A Quantitative Analysis of Instructional Designers’ Perceptions of Fairness, Accountability, and Data Privacy in Artificial Intelligence Integration

Abstract The integration of artificial intelligence (AI) into instructional design has revolutionized educational practices by enabling personalized learning and scalable solutions. However, ethical challenges such as algorithmic fairness, accountability, and data privacy remain critical concerns. This study quantitatively examined instructional designers’ perceptions of these ethical dimensions using a cross-sectional survey of 1,500 participants from diverse global contexts, including Ghana, Kenya, Nigeria, South Africa, and the USA. Three research questions guided the investigation: (1) perceptions of fairness in AI-driven learning tools, (2) accountability for AI decisions in instructional design, and (3) consideration of data privacy in AI implementation. Findings revealed significant differences in fairness perceptions by sex, with males reporting higher fairness scores than females (t(1291.51) = 8.67, p < .001, Cohen’s d = 0.456). Accountability perceptions varied by education level, with PhD holders reporting the highest scores and Diploma holders the lowest (F(4, 1495) = 13.03, p = .001, η² = .032). A multiple linear regression indicated that perceptions of fairness and accountability negatively predicted privacy considerations, though the model explained only 4.6% of the variance (R² = .046, p < .001). These results highlight the influence of demographic and educational factors on ethical perceptions and underscore the need for targeted interventions to ensure equitable and responsible AI use in instructional design.