AIEd for Good?: International Organisations, Big Tech and the Remaking of Education

• In Event: Artificial intelligence and educational policy: impact and influence

Proposal

Relevance: The rapid evolution of AI in education (AIEd) has been accompanied by considerable hype about its possibilities for enhancing teaching and learning and transforming education. AI-enhanced ‘personalised learning’, for example, – which tailors curriculum content to learners’ needs/interests – is widely promoted by both the OECD and UNESCO – two of the most prominent international organisations (IOs) focusing on education – as part of their respective visions of the future of education. This AIEd moment further includes growing emphasis on the need to ‘robot-proof’ or ‘future-proof’ learners by equipping them with ‘human-centric skills’ or ‘21st century competencies’ (OECD, 2024) such as well-being, empathy, and resilience, against the backdrop of a rapidly changing economy increasingly shaped by automation and digitisation. The proliferation of technologies, digital devices and platforms that measure and monitor these competencies helps to ensure growing emphasis on the ‘non-cognitive’ or social-emotional aspects of children’s lives, and on the ability to learn (or learnability) of human-centric capacities, alongside cognitive or academic skills and knowledge.

The specific research questions explored in this paper are as follows:

1. Who are the key actors (academics; think-tanks; policy brokers; institutions; corporations etc.) advancing AIEd and 21st century skills? Whose interests are served by AIEd advocacy or critique?
2. How is AIEd – as presented by influential educational policy actors (e.g., the OECD and UNESCO, national governments etc.) – (re)shaping understandings of education, childhood, society and what it means to be human? (Network Ethnography and Critical Discourse Analysis)

Theory/Context: Despite the growing prevalence of AI-augmented technologies in education, as a nascent technology, there is limited scholarship on how digital transformation is re-configuring our understandings of teaching and learning, of childhood, or what it means to be human. Theoretically, the research is underpinned by Sriprakash et al’s (2024) sociological framework identifying AI and related digital technologies as sociodigital practices that ‘fundamentally remake educational relationships and values: changing the way we relate to each other; what we understand learning to be; how we envisage the learner; and what we envisage as the purpose of education’ (1). Situating the field of AIEd within this broader sociodigital framework, the paper advances our understanding of the increasing involvement of private/non-state actors (e.g., tech-based philanthropists) in education and what this means for education as a public good. It builds on existing critiques of technology’s positioning as a ‘fix’ for various educational problems to illuminate vested interests and powerful networks shaping how these ‘problems’ are defined and what ‘solutions’ are being proffered. Recent analyses of policy guidance on AIEd expose their techno-solutionist and/or dystopian imaginaries (e.g., Linderoth et al., 2024). Yet our understanding of how these imaginaries evolve and take hold among the international organisations who ‘broker’ educational knowledge remains limited. Moreover, despite the emergence of 21st century skills as a ‘zeitgeist’ that has been capturing the imagination of academics, policy-makers and practitioners (Humphrey, 2013), critical interrogation of 21st century skills’ advocacy – and its evolving relationship with AIEd – is limited. This project is the first of its kind to explore the complex inter-relationships between 21st century skills and AI advocacy and their implications for education as a public good.

Mode of Inquiry

Using a mixed-methods approach combining social network analysis (SNA), network ethnography and critical discourse analysis, the paper maps relationships between key players involved in advancing or discussing AIEd globally and nationally and explores the wider educational and societal implications of how AIEd is presented by key players involved in education and AI. Firstly, based on data derived primarily from policy documents and research reports relevant to AIEd and 21st century skills, we used quantitative SNA to map relationships between key actors advancing AIEd and 21st century competencies and identify overlaps and inter-relationships between these different sets of actors (research question (RQ) 1). This allowed visualisation of the density of these networks as well as the centrality of particular actors and the power that they wield and the social relations or ‘flows’ between people and institutions involved in AIEd. This quantitative representation was augmented by a network ethnography – an approach that draws on qualitative data sources (e.g., policy documents/reports; webpages; newsletters) to illuminate the structure, dynamics and practices of policy communities and their social relationships (Ball, 2016) – to further uncover the people, organisations ideas and vested interests involved in advancing AIEd (RQ 2). Thirdly, we employed CDA to scrutinize the ideas employed by these networks and to critically consider how they are (re)shaping understandings of education, childhood, society and what it means to be human (RQ 2). Specifically, we drew on Bacchi’s ‘What is the problem represented to be?’ discourse analysis framework, which comprises six questions interrogating how particular educational problems are constituted (Bacchi & Goodwin 2016).

Findings/Implications

Our findings revealed that the AIEd policy guidance produced by IOs strongly reflects an industry logic of transforming education animated by techno-solutionism, bolstered by an over-reliance on the work of authors affiliated with companies that benefit from large-scale deployment of AI. Collectively, the analysis suggests that corporate and private interests and AI discourses are having a profound re-orienting effect on education which is in turn re-defining 21st century childhood in terms of a ‘digitally resilient learner’ imbued with ‘human-centric’ capacities and ‘transformative competencies’ such as problem-solving, persistence, empathy, self-regulation and optimism. Against the backdrop of a logic that presupposes that all learners can acquire these 21st century skills, those who are unable to become ‘digitally resilient’ are positioned as undeserving of care, rights, or justice. Our analysis further reveals that AIEd policy texts are largely silent on the fundamental concerns about ‘value extraction and industry expansion above the needs of educational communities’ Sriprakash et al. (2024, 2) or Big Tech’s ‘ubiquitous influence within and across policy streams, to primarily serve their self-interests rather than promote innovation’ (Khanal, Zhang, and Taeihagh 2014, 1). Whereas ethical considerations and concerns are not absent from the discourse of AIEd, the taken for grantedness that AI can be made ethical and promote innovation underpins the discourse.

Author

• Audrey Bryan, Dublin City University