Beyond Automation: A Systematic Review of AI Teaching Methodologies and a Framework for Human–AI Synergy in Higher Education

Artificial intelligence is increasingly present in educational settings, yet its integration remains fragmented, lacking coherent understanding of how teachers and AI can effectively collaborate. This systematic review of 169 peer-reviewed studies (2009–2024) introduces theAI-Human Synergy Framework, a model for categorizing educational AI based on autonomy and collaboration, empirically validated through systematic coding with exceptional inter-rater reliability (κ = 0.972). Our analysis reveals significant automation bias: 45.56% of implementations substitute human functions, while only 4.73% achieve genuine human-AI partnership—a 10:1 ratio. Through rigorous analysis across diverse AI methods—from Predictive Analytics to Generative AI—we demonstrate that educators are predominantly positioned as data interpreters rather than collaborative partners. While recent work (2023–2024) shows promising shifts toward Augmentation (62.5%), true Synergy remains critically underrepresented. This framework provides the first empirical quantification of the substitution-synergy gap and offers actionable implementation tools, underscoring the urgent need to prioritize collaborative design over automation as Generative AI reshapes educational practice.