What is next in the AI journey in education?

 

From Technological Adoption to Human-Centred Pedagogical Transformation

Abstract

Artificial intelligence (AI) has moved rapidly from the periphery of educational technology to a central position in contemporary debates about teaching, learning, and assessment. Early stages of AI adoption in education focused predominantly on automation, efficiency, and personalised content delivery. However, these developments have also generated significant ethical, pedagogical, and equity-related concerns, particularly regarding academic integrity, learner agency, teacher professionalism, and inclusion. This essay argues that the next phase of the AI journey in education must move beyond a narrow focus on tools and efficiency toward a systemic, human-centred, and pedagogically grounded transformation. Drawing on learning theory, critical AI studies, inclusive education research, and emerging empirical evidence, the paper examines how AI is reshaping curriculum, assessment, teacher identity, learner wellbeing, and educational governance. Attention is given to neurodiversity, universal design for learning, and the need to reframe AI as a learning partner rather than a substitute for human cognition. 

From Technological Adoption to Human-Centered Pedagogical Transformation

Artificial intelligence (AI) is no longer an emerging technology in education. Today, technologies such as large language models, adaptive learning platforms, learning analytics, and generative AI tools are widely integrated into classrooms, higher education institutions, and professional learning environments worldwide. Unlike earlier educational technologies, which primarily supported existing pedagogical practices, AI introduces a new level of disruption. It not only automates repetitive tasks but also extends into areas traditionally dominated by human expertise, including writing, problem-solving, providing feedback, and making decisions.

The current conversation about AI in education is characterised by polarisation. Supporters emphasise the unprecedented possibilities for personalisation, scalability, and enhanced learner support. In contrast, critics raise concerns about AI's potential to deskill teachers, deepen existing inequalities, and compromise the authenticity of learning experiences. These debates indicate that the critical issue is no longer whether AI will be implemented in education, but rather how it will be embedded and whose values will inform its use.

This essay explores the question: What comes next in the AI journey within education? It contends that the forthcoming phase requires a comprehensive rethinking of pedagogy, assessment, and the overall purpose of education. Instead of perceiving AI merely as a collection of tools to be managed or controlled, education systems need to cultivate approaches that are ethically grounded, inclusive, and critically informed. Such approaches should position AI as a collaborative partner in the learning process, while maintaining human agency and the importance of relational teaching.

From “AI-enhanced education” to “AI-informed pedagogy,” positioning human judgement, ethics, and relational teaching at the centre of future educational ecosystems.

1. Introduction

Artificial intelligence is no longer an emergent or speculative technology within education. Large language models, adaptive learning platforms, learning analytics, and generative AI tools are now embedded in classrooms, universities, and professional learning environments worldwide. While early educational technologies often supplemented existing pedagogical practices, AI presents a more disruptive challenge, not only automating routine tasks but also encroaching upon traditional human domains such as writing, problem-solving, feedback, and decision-making (Luckin et al., 2016; Selwyn, 2019).

Much of the current discourse surrounding AI in education is polarised. On one hand, proponents highlight unprecedented opportunities for personalisation, scalability, and learner support (Holmes et al., 2022). On the other hand, critics warn of deskilling teachers, exacerbating inequities, and undermining authentic learning (Biesta, 2022; Williamson & Eynon, 2020). These tensions suggest that the central question is no longer whether AI will be used in education, but rather how it will be integrated and whose values will shape its deployment.

This essay addresses the question: What is next in the AI journey in education? It argues that the next phase requires a fundamental reimagining of pedagogy, assessment, and educational purpose. Rather than viewing AI as a set of tools to be managed or controlled, education systems must develop ethically grounded, inclusive, and critically informed approaches that position AI as a partner in learning while preserving human agency and relational teaching.

---

2. From Automation to Augmentation: Reframing AI’s Educational Role

2.1 Early Stages of AI in Education

Initial uses of AI in education focused largely on automation and efficiency. Intelligent tutoring systems, automated grading, and adaptive content delivery were designed to replicate aspects of teacher instruction at scale (Anderson et al., 1995). These systems aligned closely with behaviourist and cognitivist models of learning, emphasising mastery, repetition, and performance optimisation.

While such approaches demonstrated measurable gains in specific domains, they also reinforced narrow conceptions of learning as content acquisition rather than meaning-making (Biesta, 2015). Furthermore, automation-driven models risked reducing learners to data points and teachers to system supervisors (Selwyn, 2019).

2.2 AI as Cognitive and Metacognitive Support

The next phase of AI integration increasingly emphasises augmentation rather than replacement. Generative AI tools, for example, can scaffold brainstorming, provide formative feedback, and model expert thinking processes (Mollick & Mollick, 2023). When used pedagogically, these tools can support metacognition, self-regulation, and reflective learning.

However, augmentation is not inherently beneficial. Without explicit pedagogical framing, AI risks becoming a cognitive crutch that undermines deep learning (Kirschner & De Bruyckere, 2017). The future of AI in education, therefore, depends on educators’ capacity to design learning experiences that make AI use visible, intentional, and critically examined.

3. Redefining the Role of the Teacher

3.1 From Knowledge Authority to Learning Architect

AI challenges traditional notions of teacher authority grounded in exclusive access to knowledge. Yet this does not diminish the role of teachers; rather, it reconfigures it. In AI-rich environments, teachers increasingly function as learning architects, ethical guides, and sense-makers (Fullan et al., 2020).

This shift aligns with constructivist and sociocultural theories of learning, which emphasise the importance of dialogue, scaffolding, and social interaction (Vygotsky, 1978). AI can support these processes, but it cannot replace the relational and contextual judgement that teachers bring to complex learning environments.

3.2 Professional Identity and Teacher Agency

A critical risk in the next phase of AI adoption is the erosion of teacher agency through algorithmic decision-making and platform-driven pedagogy (Williamson, 2017). If AI systems prescribe learning pathways, assessments, or interventions without teacher interpretation, professional judgement may be marginalised.

Sustainable AI integration, therefore, requires robust professional learning focused not only on technical skills but also on critical AI literacy, data ethics, and pedagogical decision-making (OECD, 2021). Teachers must remain active agents in shaping how AI is used, rather than passive implementers of externally designed systems.

---

4. Assessment in the Age of Artificial Intelligence

4.1 The Crisis of Traditional Assessment

Few areas of education have been more disrupted by AI than assessment. Generative AI has exposed the fragility of assessment models reliant on unsupervised written tasks and recall-based outcomes (Eaton, 2023). Attempts to “AI-proof” assessment through surveillance or detection tools have proven both ineffective and ethically problematic. This disruption, however, creates an opportunity to address long-standing critiques of assessment as reductive, inequitable, and misaligned with authentic learning (Boud & Falchikov, 2007).

4.2 Toward Authentic and Process-Oriented Assessment

The next stage of AI-informed assessment emphasises:

• Process over product
• Formative feedback over summative judgement
• Transparency over concealment of AI use

Authentic assessments, such as portfolios, oral defences, design projects, and reflective commentaries, make learning visible and value higher-order thinking (Wiggins, 1998). When AI is explicitly incorporated into assessment design, students can be evaluated on their ability to use AI critically, ethically, and creatively rather than covertly.

---

5. AI Literacy as a Foundational Educational Outcome

5.1 Beyond Technical Skills

AI literacy extends beyond knowing how to use tools. It encompasses understanding how AI systems are trained, how bias and power operate within algorithms, and how AI reshapes knowledge production (Ng et al., 2021).

The next phase of education must integrate AI literacy as a core capability alongside traditional literacy. This includes:

• Functional AI literacy (use and interaction)
• Critical AI literacy (ethics, bias, governance)
• Creative AI literacy (co-design and innovation)

5.2 Democratic and Ethical Imperatives

Without widespread AI literacy, educational AI risks reinforcing existing inequalities, as only privileged learners gain the skills to question, adapt, and shape AI systems (Noble, 2018). Embedding AI literacy within compulsory education is therefore both an educational and democratic imperative.

6. Inclusion, Neurodiversity, and Wellbeing

6.1 AI and Universal Design for Learning

AI holds significant promise for inclusive education when aligned with Universal Design for Learning (UDL) principles (CAST, 2018). Adaptive interfaces, multimodal content, and personalised pacing can reduce barriers for neurodiverse learners, including those with ADHD, autism, and dyslexia. However, inclusion is not automatic. Poorly designed AI systems may exacerbate cognitive overload, surveillance anxiety, or deficit-based profiling (Cukurova et al., 2020).

6.2 Wellbeing and Cognitive Load

The next phase of AI integration must prioritise learner and teacher wellbeing. Research on cognitive load theory highlights the risk of overwhelming learners with excessive information and choices (Sweller et al., 2019). AI systems should therefore aim to reduce extraneous load and support executive functioning, not intensify performance pressures.

Human-centred AI design, co-created with diverse learners, is essential to ensure that efficiency does not come at the cost of well-being.

---

7. Governance, Ethics, and System-Level Change

7.1 From Policy to Practice

Ethical AI in education cannot be addressed solely through high-level policy statements. It requires translation into everyday pedagogical decisions, assessment practices, and institutional cultures (Floridi et al., 2018).

Key ethical considerations include:

• Data privacy and consent
• Transparency of algorithms
• Accountability for AI-driven decisions
• Clear boundaries around appropriate AI use

7.2 Systemic Transformation

AI will increasingly shape curriculum design, resource allocation, and early intervention systems through learning analytics and predictive modelling. While these developments offer opportunities for equity-focused support, they also risk entrenching deficit narratives if not critically examined (Williamson & Eynon, 2020). The next phase of the AI journey, therefore, requires systemic governance structures that balance innovation with care, efficiency with justice, and data with human insight.

8. Conclusion

The next stage of the AI journey in education is not defined by more powerful algorithms or faster adoption. Rather, it is characterised by a conceptual shift: from AI as a technological solution to AI as a pedagogical, ethical, and human challenge.

This article argued that future-focused education systems must move beyond instrumental uses of AI toward AI-informed pedagogy grounded in inclusion, relational teaching, and critical literacy. Teachers remain central as designers of learning, interpreters of data, and guardians of educational values. Assessment must be reimagined to prioritise authentic learning and transparency. AI literacy must become a foundational outcome, and well-being must be positioned as a core design principle rather than an afterthought. Ultimately, the measure of success in the next phase of AI in education will not be how intelligently machines perform, but how thoughtfully humans learn, teach, and live alongside them.

References

Anderson, J. R., Corbett, A. T., Koedinger, K. R., & Pelletier, R. (1995). Cognitive tutors: Lessons learned. The Journal of the Learning Sciences, 4(2), 167–207.
Biesta, G. (2015). Good education in an age of measurement. Routledge.
Biesta, G. (2022). Why educational research should not just solve problems, but should cause them as well. British Educational Research Journal, 48(1), 1–4.
Boud, D., & Falchikov, N. (2007). Rethinking assessment in higher education. Routledge.
CAST. (2018). Universal design for learning guidelines version 2.2.
Cukurova, M., Luckin, R., & Holmes, W. (2020). Artificial intelligence in education: The three grand challenges. British Journal of Educational Technology, 51(6), 2143–2157.
Eaton, S. E. (2023). Postplagiarism: Transcending the binary of plagiarism and integrity. International Journal for Educational Integrity, 19(1).
Floridi, L., et al. (2018). AI4People—An ethical framework for a good AI society. Minds and Machines, 28(4), 689–707.
Fullan, M., Quinn, J., Drummy, M., & Gardner, M. (2020). Education reimagined: The future of learning. New Pedagogies for Deep Learning.
Holmes, W., Bialik, M., & Fadel, C. (2022). Artificial intelligence in education: Promises and implications for teaching and learning. Center for Curriculum Redesign.
Kirschner, P. A., & De Bruyckere, P. (2017). The myths of the digital native and the multitasker. Teaching and Teacher Education, 67, 135–142.
Luckin, R., Holmes, W., Griffiths, M., & Forcier, L. B. (2016). Intelligence unleashed: An argument for AI in education. Pearson.
Mollick, E., & Mollick, L. (2023). Using AI to implement effective teaching strategies. Harvard Business Publishing Education.
Ng, D. T. K., et al. (2021). AI literacy: Definition, teaching, evaluation and ethical issues. Computers and Education: Artificial Intelligence, 2.
Noble, S. U. (2018). Algorithms of oppression. NYU Press.
OECD. (2021). AI in education: Challenges and opportunities. OECD Publishing.
Selwyn, N. (2019). Should robots replace teachers?. Polity Press.
Sweller, J., Ayres, P., & Kalyuga, S. (2019). Cognitive load theory. Springer.
Vygotsky, L. S. (1978). Mind in society. Harvard University Press.
Wiggins, G. (1998). Educative assessment. Jossey-Bass.
Williamson, B. (2017). Big data in education. Sage.
Williamson, B., & Eynon, R. (2020). Historical threads, missing links, and future directions in AI in education. Learning, Media and Technology, 45(3), 223–235.