In the age of generative AI, we face a major challenge: the growing gap between technological advancements and a humanistic understanding of education. This divide threatens our ability to use AI tools wisely and our capacity to foresee their societal impact. Generative AI is profoundly shaping our everyday educational experiences in classrooms and in decision-making institutions. It is transforming, with remarkable speed, how we learn and create.
In traditional education, most students, regardless of background, ability or temperament, followed a general curriculum designed for the collective rather than the individual. In an era where individuality and personalisation have become pillars of modernity, the limitations of this approach are clear. Yet, many educational stakeholders – students, teachers and policymakers – struggle to grasp how generative AI can enhance individual learning while addressing ethical and societal challenges. This lack of understanding creates tensions, hindering the harmonious integration of generative AI in education.
Personalised learning
It is not that the concept of personalised learning is new. In his 1762 work Emile, or On Education, Jean-Jacques Rousseau advocated for education tailored to each student’s needs and interests. More recently, educator Célestin Freinet promoted an approach that respected each child’s rhythm and curiosity. In France, these methods have remained on the margins of the education system, limited by the demands of mass education. The 1833 Guizot law, which mandated primary education, and the 1975 Haby reform, which established a unified secondary school system, sought to promote equality through uniformity. While these reforms widened access to education, they have often been criticised for neglecting the diversity of students’ talents and aptitudes. Today, generative AI presents an opportunity to address the challenges of personalised learning that traditional education struggles to overcome.
With its data analysis capabilities, generative AI promises real-time, tailored adaptation to individual needs without overburdening teachers. Using sophisticated algorithms, generative AI can analyse students’ performances, learning styles and even preferences, designing custom learning paths that adjust levels of difficulty and types of exercises as students progress.
Harvard’s tailored generative AI tutor illustrates generative AI’s ability to personalise education. Integrated into a physics course, it significantly boosted student engagement by providing real-time support and tailored feedback. However, Harvard professors demonstrated that generative AI should augment, not replace, human instruction, emphasising the distinct yet complementary strengths of both.
While AI excels in delivering personalised feedback and fostering engagement through data-driven insights, it lacks the nuanced contextual understanding and adaptability that human educators bring to the classroom, especially in nurturing critical thinking and ethical reasoning. Indeed, overreliance on AI could undermine the teacher’s role as a guide for deeper intellectual exploration. The professors advocated for comprehensive teacher training programmes that integrate ethical and pedagogical frameworks, ensuring AI serves as a tool to enhance, rather than detract from, the humanistic mission of education.
Another area of concern is generative AI’s effect on creativity. If an algorithm guides every aspect of a student’s learning, are they still free to explore, make mistakes and pursue unpredictable paths that are often the most intellectually fruitful? Research conducted at the University of South Carolina found that while tools like ChatGPT helped students brainstorm effectively, they also made students overly reliant on generative AI, reducing their confidence in their own creative capabilities. Many students reported that generative AI’s ideas influenced their thinking, limiting independent exploration.
Teacher training
Digital-native students intuitively use these technologies, yet they often lack an understanding of the ethical and philosophical implications. Today’s teachers are caught between the call to innovate and a lack of sufficient training. To bridge these gaps, a deep rethinking of education is needed.
It is crucial to integrate generative AI epistemology into teacher training to help teachers understand how generative AI systems acquire, process and generate knowledge. For example, in France, the AI4T (Artificial Intelligence for and by teachers) project equips educators with tools such as MOOCs (massive open online courses) and open textbooks to integrate AI into classrooms.
The initiative emphasises ethical considerations like transparency and equity while fostering critical understanding of AI’s capabilities. By providing practical and epistemological training, AI4T helps teachers navigate the challenges of personalised, inclusive learning environments. Similarly, in the United States, the EducateAI initiative, launched by the National Science Foundation, provides resources for teachers across educational levels to ensure accessible and inclusive AI education. Additionally, the AI for Education organisation offers “Train-the-Trainer” programmes, enabling school staff to develop expertise in generative AI and deliver high-quality professional development within their institutions.
This training should not turn teachers into engineers but should give them insights into the ethical, social and philosophical aspects surrounding generative AI. Teachers with this background would be able to make these complex technologies more accessible to students and foster critical thinking about generative AI’s uses. This expanded role for teachers is key to democratising understanding of generative AI and encouraging an informed debate about its role in education.
Generative AI integration must not come at the expense of fostering critical thinking, creativity, empathy and the development of ethical reasoning – on the contrary, it should reinforce them. These principles, central to a humanistic understanding of education, ensure that learning remains focused on the holistic growth of individuals rather than solely on technological efficiency.
Generative AI in education should be guided by goals and values collectively defined by all educational stakeholders. It is essential to prevent these technologies from evolving autonomously, disconnected from the real needs of learners and teachers. Only in this way can we shape a future where generative AI enhances our humanity, realising a vision of technology that serves emancipatory education.