When Engagement Is Not Learning
Cognitive Safety by Design in Generative AI Tutoring
Parole chiave:
generative AI, learner agency, cognitive safety, cognitive engagement, scaffoldingAbstract
Generative AI can sustain fluent tutoring interactions while performing the reasoning learners are expected to develop. This conceptual-methodological paper develops Cognitive Safety by Design through an implementation example (Proxima ZSP) and a descriptive synthesis of published synthetic evidence. Five requirements connect assistance to inspectable evidence: diagnose before intervening, provide minimum sufficient mediation, preserve learner-owned reasoning, fade assistance and verify learning after assistance. Across 1,100 dialogues from the Synthetic Learner Evaluation Framework (SLEF), an automated reader classified 85.7–87.8% of classifiable sessions assigned low metacognition as high. The share of sessions in which transfer could be classified ranged from 0.9% to 37.7%. These results motivate explicit records of the opportunities available in each dialogue. Opportunity-based evaluation distinguishes an absent opportunity, ambiguous evidence and an interpretable response, preventing missing evidence from automatically becoming a judgment of low capability. The developing L×M×C framework—depth of reasoning (L), expressed metacognition (M), and consolidation and transfer (C)—provides provisional outcome hypotheses, including session and spacing criteria. The contribution makes attribution explicit: each claim about a learner is linked to the assistance received, opportunities available and subsequent independent performance.
Riferimenti bibliografici
Aleven, V., McLaren, B. M., Roll, I., & Koedinger, K. R. (2006). Toward meta-cognitive tutoring: A model of help seeking with a Cognitive Tutor. International Journal of Artificial Intelligence in Education, 16(2), 101–128. https://doi.org/10.3233/IRG-2006-16(2)02
Bastani, H., Bastani, O., Sungu, A., Ge, H., Kabakcı, Ö., & Mariman, R. (2025). Generative AI without guardrails can harm learning: Evidence from high school mathematics. Proceedings of the National Academy of Sciences, 122(26), e2422633122. https://doi.org/10.1073/pnas.2422633122
Cen, H., Koedinger, K. R., & Junker, B. (2006). Learning factors analysis—A general method for cognitive model evaluation and improvement. In M. Ikeda, K. D. Ashley, & T.-W. Chan (Eds.), Intelligent tutoring systems (pp. 164–175). Springer. https://doi.org/10.1007/11774303_17
Chi, M. T. H., & Wylie, R. (2014). The ICAP framework: Linking cognitive engagement to active learning outcomes. Educational Psychologist, 49(4), 219–243. https://doi.org/10.1080/00461520.2014.965823
Fan, Y., Tang, L., Le, H., Shen, K., Tan, S., Zhao, Y., Shen, Y., Li, X., & Gašević, D. (2025). Beware of metacognitive laziness: Effects of generative artificial intelligence on learning motivation, processes, and performance. British Journal of Educational Technology, 56(2), 489–530. https://doi.org/10.1111/bjet.13544
Fredricks, J. A., Blumenfeld, P. C., & Paris, A. H. (2004). School engagement: Potential of the concept, state of the evidence. Review of Educational Research, 74(1), 59–109. https://doi.org/10.3102/00346543074001059
Iannacone, M. (2026a). A multidimensional framework for evaluating reasoning in AI-mediated maieutic tutoring: Theoretical foundations and operational specification (Version 1.2.7) [Preprint]. EdArXiv. https://doi.org/10.35542/osf.io/hvx37_v2
Iannacone, M. (2026b). Can assigned learner states be recovered from synthetic tutoring dialogues? A factorial study with SLEF [Preprint]. EdArXiv. https://doi.org/10.35542/osf.io/esmx2_v1
Iannacone, M. (2026c). Preserving learner agency in artificial intelligence-mediated education: An open framework for evaluating learning, metacognition and cognition [Manuscript prepared for an accepted ICERI2026 contribution].
Iannacone, M. (2026d). SLEF Reference Evaluation Dataset (Version 1.1.0) [Data set and software]. Zenodo. https://doi.org/10.5281/zenodo.23068043
Kapur, M. (2008). Productive failure. Cognition and Instruction, 26(3), 379–424. https://doi.org/10.1080/07370000802212669
Kestin, G., Miller, K., Klales, A., Milbourne, T., & Ponti, G. (2025). AI tutoring outperforms in-class active learning: An RCT introducing a novel research-based design in an authentic educational setting. Scientific Reports, 15, 17458. https://doi.org/10.1038/s41598-025-97652-6
Koedinger, K. R., & Aleven, V. (2007). Exploring the assistance dilemma in experiments with Cognitive Tutors. Educational Psychology Review, 19(3), 239–264. https://doi.org/10.1007/s10648-007-9049-0
Maurya, K. K., Srivatsa, K. A., Petukhova, K., & Kochmar, E. (2025). Unifying AI tutor evaluation: An evaluation taxonomy for pedagogical ability assessment of LLM-powered AI tutors. In Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers) (pp. 1234–1251). Association for Computational Linguistics. https://doi.org/10.18653/v1/2025.naacl-long.57
Mislevy, R. J., Almond, R. G., & Lukas, J. F. (2003). A brief introduction to evidence-centered design. ETS Research Report Series, 2003(1), i–29. https://doi.org/10.1002/j.2333-8504.2003.tb01908.x
Risko, E. F., & Gilbert, S. J. (2016). Cognitive offloading. Trends in Cognitive Sciences, 20(9), 676–688. https://doi.org/10.1016/j.tics.2016.07.002
Scarlatos, A., Lee, J., Woodhead, S., & Lan, A. (2026). Simulated students in tutoring dialogues: Substance or illusion? In Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) (pp. 42349–42385). Association for Computational Linguistics. https://doi.org/10.18653/v1/2026.acl-long.1960
Soderstrom, N. C., & Bjork, R. A. (2015). Learning versus performance: An integrative review. Perspectives on Psychological Science, 10(2), 176–199. https://doi.org/10.1177/1745691615569000
Sweller, J. (1988). Cognitive load during problem solving: Effects on learning. Cognitive Science, 12(2), 257–285. https://doi.org/10.1207/s15516709cog1202_4
van de Pol, J., Volman, M., & Beishuizen, J. (2010). Scaffolding in teacher-student interaction: A decade of research. Educational Psychology Review, 22(3), 271–296. https://doi.org/10.1007/s10648-010-9127-6
van de Pol, J., Volman, M., Oort, F., & Beishuizen, J. (2014). Teacher scaffolding in small-group work: An intervention study. Journal of the Learning Sciences, 23(4), 600–650. https://doi.org/10.1080/10508406.2013.805300
VanLehn, K. (2011). The relative effectiveness of human tutoring, intelligent tutoring systems, and other tutoring systems. Educational Psychologist, 46(4), 197–221. https://doi.org/10.1080/00461520.2011.611369
Vygotsky, L. S. (1978). Mind in society: The development of higher psychological processes. Harvard University Press.
Wood, D., Bruner, J. S., & Ross, G. (1976). The role of tutoring in problem solving. Journal of Child Psychology and Psychiatry, 17(2), 89–100. https://doi.org/10.1111/j.1469-7610.1976.tb00381.x
Yao, K. (2026). What does the credential still certify? Cognitive stewardship for AI-mediated education [Preprint]. arXiv. https://doi.org/10.48550/arXiv.2607.19988
Zainuddin, Z., Churiyah, M., Xiaoyu, W., & Halili, S. H. (2026). From AI literacy to cognitive AI safety: Rethinking safe AI use in education. Interactive Technology and Smart Education. Advance online publication. https://doi.org/10.1108/ITSE-12-2025-0396
##submission.downloads##
Pubblicato
Come citare
Fascicolo
Sezione
Licenza
Copyright (c) 2026 Marco Iannacone

Questo lavoro è fornito con la licenza Creative Commons Attribuzione - Non commerciale - Non opere derivate 4.0 Internazionale.