Early Insights from our Ministries Challenge
Three months into AI Observatory’s Ministries Challenge
I’m motivated by solving our own problem”, says Shefat Islam, a Research Officer at the Directorate of Secondary and Higher Education’s (DSHE) Research and Innovation Division in Bangladesh, “by finding out what are the root causes of students dropping out [of school] at secondary level. I want to know, how can we support our teachers with the assessment of students? How can we use this assessment data to better support students, to stop dropouts? And how can we support policy makers to find out the real issues in our schools?”
As Shefat highlights, Ministries of Education around the world face challenges at multiple, complex levels. Decades of donor-funded education systems around the world means that countries have been put under “significant pressures to collect education data”.
Ministries are rising to the challenge
The scale of the challenge now faced by low- and middle-income countries is now significant. As aid cuts come into play globally, and information amounting to tens of millions of data points per country is collected — from school maintenance through the student exam results through to child health records — Ministries of Education are having to consider how to do more with less.
The AI Observatory launched the Ministries of Education AI Challenge earlier this year, to support ministries to address this reality. We are working with governments to explore practical applications of AI in education service delivery, assess risks and opportunities, and share evidence on what works to support more equitable learning outcomes.
Our hypothesis is: if ministries are given adequate support to test AI for real-world delivery challenges, then they will build lasting capabilities to use AI responsibly in order to achieve more with less.
Across six countries, there is common ground
Upon launching the Challenge, we received over 30 applications from 16 countries, and we are now working with Ministry teams in the Philippines, Pakistan, Bangladesh, Kenya, Nigeria and Sierra Leone.
Each context brings its own challenge: from the scale of the primary education system in Nigeria, to overseeing an education system scattered across hundreds of islands in the Philippines. Yet, there are several similar issues that each of these ministries wants to address:
- School maintenance logs – how can early warning systems notify ministries about issues such as building maintenance, and upgrades to infrastructure, particularly where schools are oversubscribed.
- Student dropouts – how can AI develop early warning systems to flag falling student attendance rates, highlight common issues shared by these students and therefore support policy makers in addressing these challenges?
- Learning pathways – where students are struggling with subjects, how can AI identify areas to work on and set up appropriate resources to support teachers, such as generating customised lesson plans?
Our latest webinar and blog shares more from Ministries’ perspectives about the issues they are aiming to tackle.
Blending expert guidance with peer learning supports AI adoption
Participating ministries receive tailored support and form dedicated teams that have access to shared learning through a collaborative platform and tailored events, with each team comprising:
- Ministry staff who are supported within their roles to commit dedicated time to the Challenge.
- AI experts who work as individuals or in teams to integrate with the ministry team and support with technical know-how.
- Country education specialists who support with navigating systems and stakeholders in Ministries of Education and beyond in the wider government space.
- Coaches who offer hands-on support around innovation approaches and methods.
The final piece of the puzzle is the shared learning experiences offered through the AI Observatory and Action Lab.
By partnering with Apolitical’s Government AI Campus, we have been able to expedite registration for ministry staff or affiliated partners. This has been hugely popular with the Federal Ministry of Education team in Nigeria, who are invested in the capacity building opportunities offered through the Ministries of Education AI Challenge.
AI expert, Essa Mohamedali, has been running interactive learning sessions and developing AI resources for all teams. The team in Bangladesh, which is made up of colleagues from various agencies, has requested that Essa run online workshops to build capacity in colleagues from across the AI adoption space in Bangladesh, showing that this hands-on support to the cohort of Ministries from a dedicated AI expert is useful for Ministries to quality assure and check their own AI explorations.
Expert partnerships deliver quick wins that build long-term confidence in the power of AI
In Sierra Leone, a Director at the Ministry of Basic and Senior Secondary Education has been working with the AI Expert, Mohamed Alpha. Together, the Director and Mohamed have been able to use AI to identify eight data points and longitudinal analysis that could help the team allocate resources across Ministry sections and geographies.
After just one meeting, Mohamed was able to use AI to generate a report for the Director, including all eight data points, in-depth analysis and policy recommendations. This quick win not only allowed the Director to lead the team on evidence-based decision making at the Ministry, but also develop a deeper understanding of AI’s capabilities.
In the Philippines, the E-CAIR (Education Center for AI Research) team has moved around different government departments; starting as CAIR in the Philippine Department of Trade and Industry and moving to become E-CAIR in February 2025. The team’s challenge is focussed on how they can test and validate this model of a centre, sitting within government.
Through mapping exercises with their innovation coach, Gita Luz, the team now recognise how busy and fragmented the AI space is at the national level, with the Department of Science and Technology leading on the AI roadmap while individual agencies pursue their own AI policies and programmes. Gita has been working with the team to identify and outline innovation pathways – how to plan, test, and validate an approach so that the centre has cross-government buy–in and longevity. This enables E-CAIR to consider how to position themselves within a crowded government landscape and develop a roadmap for influencing this position.
Working together to integrate ministry know-how, AI technical support, innovation expertise, country and context knowledge, teams are starting to learn about which AI approaches can have real impact in education service delivery.“This is what we want to do on this pilot project,” says Shefat Islam, “build a prototype solution that starts to answer our questions.”
The Ministries of Education AI Challenge runs from July 2025 to March 2026. Follow along with our learnings by signing up to the EdTech Hub newsletter and selecting “AI”.
The AI Observatory is made possible by support from UK International Development. Please follow along and join the conversation on LinkedIn!
