In recent years, the world has witnessed a surge of advancement and popularization of artificial intelligence (AI) technologies. With the publication of several UNESCO (2019; 2021; 2023) documents and a range of studies positioning AI as a key point of innovation and renovation of the enterprise of education, the rise of ‘AI in education’ (AIED) seems inevitable.

Not only is there a general trend of people increasingly trying to apply AI technologies to the organization and practice of teaching and learning, but there also is a (rather urgent) call for educators and school administrators to learn about AI and implement AI in their educational practices. It seems that we are expected nothing less than to ‘embrace’ (Walter, 2024) this trend both because it is conducive to improving education (whatever this might mean) and because it is crucial for our own survival as participants of the contemporary educational life-world.

However, we, the authors of this blog post, cannot but feel an urge to be contradictorians against this narrative of what we would like to call the ‘AI gospel’. We borrow this phrase from the classic sociological work of Grubb and Lazerson (2004) who posited the idea of the ‘education gospel’ to satirize the narrative that education could and should be a panacea for many social (and possibly also individual) ailments. ‘AI gospel’ reflects a similar point – that is, there is now a narrative about AIED that the many problems with contemporary education could be resolved or at least alleviated to a large extent by the introduction of AI technologies and AI literacy to classrooms (for teachers and students) and households (for parents and children).

There are at least three AI gospels that permeate the contemporary educational discourse. The first and perhaps the most popular gospel regarding AIED is that AI technologies could provide novel and effective ways to facilitate learning such that learning is made easier, more convenient, more personalized, and less reliant on ‘traditional’ elements of education such as teachers, schools, and parents. This seems particularly promising because it fits well with the prevalent learner-centred and learning-centred discourse. However, as Gert Biesta (2006) quite presciently points out, this ‘learnification’ of education – the reduction of education to the facilitation of learning – comes at the cost of diminishing the possibility for teacher and teaching to offer students something beyond the latter’s current capacity and interest and to point them towards something that might resist rather than gratify their immediate desires.

The second gospel is the promise of AIED to alleviate educational inequalities that are assumed to be attributable to the lack of access to ‘quality’ educational resources. AI, manifesting in the form of standardized programs, could offer users an equal opportunity to attain ‘quality’ education that fits each of them (rather than any given agenda, institution, or culture that might work against their welfare). This gospel could be doubted firstly because such a promise has already been promulgated since the rise of the internet. It is quite obvious now that the mere guarantee of equal access does not guarantee the equality of opportunity to succeed. On a more fundamental level, this gospel is problematic because it envisions an education that prizes depersonalization (treating students as sets of programmatic data) through the narrative of personalization and standardization (treating students as different but entirely replaceable learning subjects) through the narrative of equality.

One gospel that targets teachers and parents in particular is the narrative that AI technologies would only empower but never replace the unique and indispensable roles that teachers and parents could play in a given student’s or child’s course of development and becoming. The lure of this promise lies in the fact that AI could indeed help teachers and parents to clear away much of their work that requires redundant manual labour, which, as the narrative goes, would allow educators to have more freedom to engage in the more creative and relational aspects of their work. However, we must be vigilant about the possibility that some of what seems redundant and replaceable by AI is actually essential to the work of education and educators. It may also be worth-contemplating whether those aspects of education that are replaceable by AI are indeed necessary for or conducive to the work of education.

So, what is wrong with the AI gospel? A gospel is just what it is – an article of faith that is filled with so much optimism and ‘shining light’ that it blinds alternative viewpoints and distracts people from attending to the potential dangers of what it preaches. The AI gospel could indeed have many grains of truth, but it could also be used as a hegemonic narrative to force political, financial, and even personal agendas onto educators and educands while simultaneously proclaiming that it is all for their own benefit.

Reference List

Biesta, G. (2006). Beyond learning: Democratic education for a human future. London: Routledge.

Grubb, W. and Lazerson, M. (2004). The education gospel: The economic power of schooling. Cambridge: Harvard University Press.

UNESCO (United Nations Educational, Scientific, and Cultural Organization) (2019). Beijing Consensus on artificial intelligence and education. Available at: https://unesdoc.unesco.org/ark:/48223/pf0000368303 (Accessed 22 January 2025).

UNESCO (United Nations Educational, Scientific, and Cultural Organization) (2021). AI and education: guidance for policy-makers. Available at: https://unesdoc.unesco.org/ark:/48223/pf0000376709 (Accessed 15 May 2025).

UNESCO (United Nations Educational, Scientific, and Cultural Organization) (2023). Guidance for generative AI in education and research. Available at: https://unesdoc.unesco.org/ark:/48223/pf0000386693 (Accessed 24 May 2025).

Walter, Y. (2024). Embracing the future of artificial intelligence in the classroom: The relevance of AI literacy, prompt engineering, and critical thinking in modern education, International Journal of Educational Technology in Higher Education, 21 (1), 15-29.