UNESCO International Day of Education 2025: Artificial Intelligence and Education
Mr Xing Qu, Deputy Director-General of UNESCO opening the International Day of Education at UNESCO, Paris © Heritage Education Network Belize
On the International Day of Education (February 24th 2025), UNESCO organised an event on Artificial intelligence and education. We attended. here are some of the highlights.
*Please note that the below are excerpts of personal notes taken during dialogues, which may lack some statements and contain paraphrasing and personal bias. You may re-watch the live recording of the sessions here. For links to UNESCO publications and resources on AI in Education scroll to the bottom.
Ideas in dialogue AI and education: Preserving human agency in a world of automation
Panel on Ideas in dialogue AI and education: Preserving human agency in a world of automation, UNESCO, Paris © Heritage Education Network Belize
Speakers:
• Mr François Taddei, Chief Exploration Officer, Learning Planet Institute (Evolutionary Biologist)
• Ms Payal Arora, Professor of Inclusive AI Cultures, Media and Culture Studies Department, Utrecht University, Kingdom of the Netherlands (online) (Anthropologist)
Key Thoughts
1) AI is dynamic and used to be at an infant's level, then a student’s level, and now at a young researcher's level. It brings rapid changes in education and learning. We need a more dynamic system where everyone adapts to these changes. For this, we need to understand how AI works and know how to look at it.
2) People need to develop awareness of their agency, as it can be amplified by AI. For example, if a bias is present, it can be amplified by AI. An example was highlighted of a project where the focus is on humans NOT asking questions from AI but having AI ask questions from humans to come to conclusions and generate suggestions. This allows different approaches to building and guiding knowledge in a more open way.
3) There is a lot of anxiety around AI replacing teachers. However, AI is simply a system embedded in a much larger process of education as a whole. Learning happens in all aspects of life and socio-emotional support and connections are also necessary. Therefore, human teachers cannot be replaced. The questions we can ask are not how teachers will be replaced but rather how AI can help teachers.
4) Education has been exclusionist and not inclusive to begin with. Now that AI allows for information access, the “classroom” has been sidelined. Many girls and women or other chronically discriminated people who are not allowed access to the public sphere of education now have another tool.
In addition, the classroom is not the most ideal place for self-actualisation. Many people are not allowed or have the ability to participate in forms of classroom education. There are also exclusions and restrictions in curriculum, like sex education, gender studies or exclusion of histories. AI is a tool that allows us to open our minds and access information. That said, we have our work cut out for us as AI can also grow biases.
Drawing on this, public education should not be only happening in the classroom. The classroom is one of the spaces, but we need to enable learning as a whole.
Discussion on Ethics
1) There is a need for balance in co-evolving intelligence and ethics. It is not a new concept special to AI, as in the past there have been intelligent people who have not been very ethical as well.
2) AI is shaped by the datasets it is trained on. It is reflective of the material it is fed. As an example, in a project, a group identified landmarks to be visible on a map. When building the back-end data sets, they asked what was important to them. In their context, water tanks were viewed as an important landmark as their communities relied on these tanks. Would they be viewed as landmarks by other groups? It is important to build diverse datasets from the start to eliminate the biases, however, it is an enormous task.
AI integration in schools: Lessons from early adopters
Panel on AI integration in schools: Lessons from early adopters, UNESCO, Paris © Heritage Education Network Belize
Speakers: Speakers in this session were principals and heads of schools which are early adopters of AI technologies.
• Mr Antoine Delaitre, Head of International School of Paris, France
• Mr John Dalton, Co-Principal of David Game College, United Kingdom
• Mr Michel Muvange Nturo, Headmaster, Kayonza Modern Secondary, Rwanda (online)
Key Thoughts:
1) The schooling landscape is highly unequal. Some schools banned the use of mobile technologies in class and some clubs reject technology. AI has been used to enhance education but has also been used as a weapon to destroy education. Some schools have no access to AI or the internet, while other schools are early adopters.
2) In the International School of Paris in France, the school surveyed how people view AI, while the primary goal was to enhance learning. Teachers led the program even though a large number of them were not familiar with AI yet. This turned out to be an interesting and exciting challenge. It was a co-creation process where they took into account the views and concerns of teachers and built on them, while also learning how to bring teachers on board in the process.
3) In the David Game College in the United Kingdom, classes are taught purely by AI. There are no humans involved. Students take three hours of around 20-minute intervals of classes. Then, they focus on soft skills like Socratic reasoning, financial intelligence etc. The school adopted/created personalised learning with an adaptive system. It requires the students to learn mastery before they can move on and also allows them to revisit past curricula if a student is not yet proficient. They found that all students were learning faster than in a traditional class, for example, a girl with an average IQ learned English in 4 months.
4) In Kayonza Modern Secondary in Rwanda, they have continuous professional development sessions every week. They implemented artificial intelligence in learning at the school level. A teacher simply used ChatGPT to enhance her teaching methods.
Discussions on the Learner's response
International School of Paris, France.
2) Drawing on the UNESCO AI Competency framework for students (linked below). The school went back to the basic guiding principles of their institution and simply combined them with the statements and guidelines from the framework document.
3) The school also used AI to unpack the competency framework and get guidance on how to change lessons, methods etc., as different schools have different purposes. Some schools aim for high test scores, others aim to deliver and improve learning outcomes, and others focus on student support and community. These factors change how to use AI to achieve the specific goals.
David Game College, United Kingdom
1) There was an initial resistance to AI from the parents and students.
2) As the school is international, they take on students who were taught many different curricula. Commonly, a teacher would assume students know a topic when in reality, it was never discussed previously in their country's curriculum. The school’s system identifies what students do not know without judgment and focuses on motivating students to learn. For example, when some students do not understand the basics of mathematics, they can be discouraged and say they hate maths or that they are not good at math. This is a closed mindset. The system allows them to go back 5 years in the curriculum and re-learn even their times tables again without judgement, master the skills and then move on. This allows students to realise that they are not that bad at mathematics after all.
3) The system addresses emotional aspects and special educational needs; as bright students can feel bored and want to accelerate their learning, the school’s system can let them learn ahead. Similarly, students who have been bullied or previously had negative experiences can also have an adapted curriculum. Generally, the point is to bring the love of learning back!
Preserving human agency in an AI era: Stakeholder responses
Panel on Preserving human agency in an AI era: Stakeholder responses, UNESCO, Paris © Heritage Education Network Belize
Speakers:
• Ms Farida Shaheed, United Nations Special Rapporteur on the Right to Education (online)
• Ms Marie-Caroline Missir, Director General, Réseau Canopé, France
• Ms Laura Gregory, Senior Education Specialist and Global Lead on Teachers, World Bank
• Mr Ryan Imbriale, Vice President and Head of Worldwide Education, PowerSchool and EDSAFE AI Industry Council member
Key Thoughts:
1) Ms Farida Shaheed spoke about the risks of AI and automated tools in education, with regards to discrimination, increased digital divides, biases, questions of ethics and human rights, and challenged previously discussed topics on AI leading education and lessons.
At least 1/3 of the world is not digitally connected and as a result, AI is exaggerating inequalities.
Young people do not distinguish between "on and off [technology]" life. It is just life.
Quality education requires quality educators.
Education includes cognitive, socio-empathetic, and social development. This cannot be replaced by AI and automated tools. We should invest in educators.
The impact of Ed-tech needs to be vigorously revised and monitored.
AI surveillance in Education must be banned.
AI is pushing us into a monoculture.
Standardised tests requiring automated decisions are challenging. Replies to standardised test questions can vary based on the different realities a student may observe in their environment. Should they reply with a standard response that gets them the point or should they reply based on their observation?
We need to allow for unusual questions and prioritise community-driven innovation.
2) Teachers signing up for sessions on AI are primarily interested in how to support their work as teachers and how to help students understand AI (for example formulate prompts).
3) Ms Laura Gregory mentioned that one of the main concerns at the World Bank is the digital divide. Low-income economies could be excluded, exasperating inequalities even further.
4) Deep learning happens when the learner experiences the world and builds on pre-existing knowledge. We still learn the same way, even with the rise of AI. To learn a concept we need lived experiences to interact with others, and build relationships. However, AI can shortcut a lot of these learning activities.
5) Humans should override AI decisions to ensure that human values and ethics are prioritised.
6) The World Bank is looking carefully at AI to protect children and all people but also to ensure quality learning for all.
7) Approximately 1.5 trillion dollars were invested into AI, even though AI’s return on investment is not as clear as the return on investment in Education. It is unlikely that this investment is going to be re-directed into Education, however, we can look at how AI can support Education.
8) Mr Ryan Imbriale, Vice President and Head of Worldwide Education, PowerSchool and EDSAFE AI Industry Council member stated that they are using AI in areas where teachers are already comfortable. At PowerSchool, they are adopting a human-centred approach and stand their ground that a human should always have the ability to interact with and override AI. A teacher should always have the ability to go back and change prompts, questions etc. to ensure that the experience is human-led. Similarly, transparency in the work they are doing with AI is crucial. Community and educators should understand what is happening behind the scenes.
UNESCO & UN Resources on AI
- AI competency framework for teachers: https://www.unesco.org/en/articles/ai-competency-framework-teachers
- AI competency framework for students: https://www.unesco.org/en/articles/ai-competency-framework-students
- Guidance for generative AI in education and research: https://www.unesco.org/en/articles/guidance-generative-ai-education-and-research
- AI and education: guidance for policy-makers: https://unesdoc.unesco.org/ark:/48223/pf0000376709
- Generative AI and the future of education: https://unesdoc.unesco.org/ark:/48223/pf0000385877
- Bridging the green and digital transitions through education: https://unesdoc.unesco.org/ark:/48223/pf0000391873
- Education in the age of artificial intelligence: https://courier.unesco.org/en/articles/education-age-artificial-intelligence
- Artificial intelligence, blockchain and extended reality in lifelong learning: https://unesdoc.unesco.org/ark:/48223/pf0000391599
- K-12 AI curricula: a mapping of government-endorsed AI curricula: https://unesdoc.unesco.org/ark:/48223/pf0000380602
- AI and the Holocaust: rewriting history? The impact of artificial intelligence on understanding the Holocaust: https://www.unesco.org/en/articles/ai-and-holocaust-rewriting-history-impact-artificial-intelligence-understanding-holocaust
- Understanding the impact of artificial intelligence on skills development: https://unesdoc.unesco.org/ark:/48223/pf0000376162
- Beijing Consensus on Artificial Intelligence and Education: https://unesdoc.unesco.org/ark:/48223/pf0000368303
- Recommendation on the Ethics of Artificial Intelligence: https://unesdoc.unesco.org/ark:/48223/pf0000381137
- Dubai Declaration: https://unstats.un.org/sdgs/hlg/dubai-declaration/
More on UNESCO'S Artificial intelligence in education: https://www.unesco.org/en/digital-education/artificial-intelligence
Written by Ella Békési