How might we avoid dependence by establishing open, decentralised AI-powered networks for learning?
How might we avoid dependence by establishing open, decentralised AI-powered networks for learning?
In a recent strategic foresight study in partnership with IDRC, EdTech Hub observed a trend that the increase in AI-enabled education would create for-profit / ‘Big Tech’ dominance of AI products in education systems. Policymakers need to build resilience against a future in which private-sector companies have outsized influence over education systems. (Rahman & Freeman, 2025)
EdTech Hub’s AI Observatory is exploring how ministries of education can align partnerships – which means driving equitable outcomes through purposeful collaboration and collective action.
This week, in Issue No. 27 of the #WaypointWednesday, we spotlight strategic foresight methods for envisioning AI governance futures, open models, and innovative approaches to data sovereignty.
Early signals
Envisioning AI governance futures in which dependence is avoided
One way to move beyond the status quo is to envision alternative models in which people and institutions do not become reliant on a narrow set of providers.
- LMICs – Research on the influence of “Big Tech”: EdTech Hub is researching how low- and middle-income countries (LMICs), as the primary provider of public education for their citizens, can mitigate against undue influence and control through AI products from global EdTech private sector, whilst leveraging the potential benefits of their enhancements in education. The research co-generates learnings with Ministry of Education officials from various countries. (Adam, 2025)
- Global – Nine essays on achieving responsible AI: Bringing together voices from industry and government, civil society and academia, and drawing on perspectives from around the world, this collection of essays explores innovative approaches to governance of AI from open-source models to publicly owned organisations. (Chatham House, 2024)
Novel innovations to support locally owned data
Linguistic and cultural data underpin inclusive learning technologies. When such data are extracted from communities inequitably, education risks dependency on external platforms, loss of linguistic and cultural diversity, weaker accountability, and fewer locally relevant innovations.
- South Africa – The Esethu Framework: Designed to ensure equitable benefit-sharing, Lelapa AI created a novel community-centric data license in which native-speaking communities retain ownership of their linguistic data, receiving licensing fees from commercial entities. (Rajab et al., 2025)
Open models as alternatives to commercial systems
While “open” is not necessarily justice-oriented or equitable (Open at the Margins, 2020), open large language models (LLMs) are increasingly seen as viable alternatives to commercial systems, which are largely developed behind closed doors in the United States or China. (ETHZurich, 2025)
- Global – A fully open LLM: Researchers at EPFL, ETH Zurich, and the Swiss National Supercomputing Centre, are co-creating a “fully” open large language model. Backed by collaboration with NVIDIA and Hewlett Packard Enterprise, researchers highlight how a joint public-private effort can drive sovereign infrastructure, fostering open innovation. (ETHZurich, 2025)
Reflections:
- As private sector-led AI is increasingly being used in public education, we need to ensure that its adoption does not prioritise privatisation, marketisation, or datafication at the expense of education as a public good. Safeguarding relational and transformative learning purposes, and maintaining public, democratically accountable control, rather than allowing decision-making to shift toward private interests, will be key. (Wieczorek, 2025)
- Governments should guard against product lock-in and market capture, especially when partnering with Big EdTech (“free” or procured) without adequate capacity to regulate or negotiate terms. As education systems become more dependent on a platform’s ecosystem, users may become locked in, and learners may be unable to opt out of data use without exiting the public education system. (OECD, 2024)
Swipe for a quick take 👇🏽
We’d love to hear from you! What’s been shaping your thinking on AI? Drop your thoughts (and reading recommendations) in the comments. Explore more from EdTech Hub’s AI Observatory.
EdTech Hub’s AI Observatory is made possible with the support of the UK’s Foreign, Commonwealth and Development Office.
