UNESCO´s AI competency frameworks for teachers and students
(This post has been AI-translated from the original post in Spanish. Please contact me if you spot any inconsistencies at maricruzgarciavallejo@gmail.com)
UNESCO presented its AI competency framework during Digital Learning Week, held in Paris from September 2 to 5. In doing so, the organization joins the initiatives of numerous other national and international agencies that have developed frameworks to categorize and classify the set of technical knowledge, digital skills, and cognitive processes—ranging from simple to complex—often grouped under the broad and somewhat vague concept of 'AI competencies.
What does the expression 'AI competencies' mean, and what competencies are we talking about?
Although I previously explored the concept of AI competencies in an earlier post, I will provide a brief explanation here for readers who may not be familiar with the term.
The expression 'competencias en IA' is the Spanish translation of the English term 'AI literacy,' which began to appear in academic literature around 2020. Both the original English term and its Spanish translation are relatively new, having emerged after the rise of generative AI in education. In addition to 'AI competencies,' other Spanish terms like 'alfabetización en IA' (literary AI literacy) or 'capacitación en IA' (AI training) are considered equally valid in some publications.
In their publication Conceptualizing AI Literacy: An Exploratory Review, Davy Tsz Kit Ng and other authors made the first academic attempt to conceptualize the then-emerging term 'AI literacy,' exploring how to teach and assess the new knowledge and skills it encompasses. After a systematic review of academic literature, Ng and his collaborators concluded that AI literacy is an 'umbrella concept' that includes technical knowledge and skills already categorized under 'digital literacy,' but also involves complex cognitive processes such as the ability to formulate a problem properly.
What characterizes UNESCO's proposed AI competency frameworks?
In fact, UNESCO’s AI competency framework is similar to other initiatives, such as the 'Teachers' Competences Framework' proposed by the European Digital Education Hub.
The AI competency framework for students distinguishes four key competencies or dimensions:
Human-centred approach to AI: AI learning should focus on humans, not on AI itself.
AI ethics: Learning practices should be safe and incorporate ethics by design.
AI techniques and applications: Provide fundamental knowledge and skills related to artificial intelligence.
AI system design: Encourage problem-solving, creativity, and thinking around how to design complex processes.
For these four key competencies, which encompass both foundational technical knowledge and higher-order cognitive processes, such as the ability to formulate and solve a problem, three levels of progression are now established:
Progression Level One: Understand
Progression Level Two: Apply
Progression Level Three: Create
In the student framework, the definition of AI ethics at each of the three levels is much clearer than in the framework for teachers:
At the first level (Understand), students are expected to recognize ethical issues related to AI, such as bias, transparency, and data privacy.
At the second level (Apply), students should be able to apply ethical principles in practice by analyzing real-life scenarios and understanding how ethical decisions influence AI development and deployment.
At the third level (Create), students are encouraged to incorporate ethics by design, embedding ethical considerations in the process of creating or designing AI systems, ensuring that the outcomes are aligned with societal values and human rights.
The competency framework for teachers focuses on professional development and shares key competencies or dimensions with the student framework, such as:
Human-centred approach to AI,
AI ethics,
Foundational AI knowledge and skills.
In addition to these competencies shared with students, two new competencies are introduced:
AI pedagogy,
AI for professional development.
The main feature of UNESCO’s framework is that, unlike other national or international frameworks, it introduces the concept of AI pedagogy for the first time, which is explicitly defined as:
'A set of competencies required for purposeful and effective AI–pedagogy integration. This covers the ability to validate and select proper AI tools and to integrate them into pedagogical strategies to support course preparation, teaching, learning, socialization, social caring, and learning assessment.'
In other words, AI pedagogy is not defined as a new pedagogical model or a different approach to educational practice, but rather as 'the set of competencies required' for the effective integration of AI with pedagogy.
As with the student framework, the teacher competency framework distinguishes three progression levels, defined as:
Progression Level One: Acquire.
Progression Level Two: Deepen.
Progression Level Three: Create.
These three levels of progression, and the expectations for each across the five key competencies identified for teachers, are summarized in the original table within the document:
The UNESCO document provides a more detailed outline of the specifications for each level of competency and progression. Although the goal of this post is not to offer a summary of the framework (which is already quite clear and concise), I would like to highlight the following aspects about these two new frameworks:
UNESCO does not use any colour-coding that could lead to false conclusions, as is the case with the AIAS framework by Canadian Leo Furze (which I will discuss in a future post). There are no negative connotations (in red) for the acquisition level, and the three progression levels for each competency are presented as symmetrical and equally important. It is a progression, not a race.
A significant academic debate arose during the 2023-24 academic year about whether an 'AI pedagogy' truly exists or if current digital pedagogies should simply be adapted to the use of new generative AI tools. UNESCO seems to lean toward the latter, defining AI Pedagogy as the knowledge and competencies required to integrate AI into existing pedagogical practices.
There is a clear need to train both teachers and students in what is called 'AI ethics' (or 'IA ethics' in the original English text). However, in the teachers' framework, the characteristics of the ethical code that teachers should adopt at each of the three progression levels are not quite defined.
I am unsure whether this ambiguity is intentional, allowing national and supranational frameworks to define what AI ethics entails based on the moral and social codes of each country or society, or whether UNESCO plans to publish a common ethical framework for AI use in the future. I have my doubts about a 'common' framework, which could lead to a new form of cultural colonization (the UN’s global agenda—or 'globalist,' according to some—is the agenda of the countries that won World War II and is largely pro-capitalist, with a few exceptions such as Russia). Each country should adopt its own ethical framework for AI use, especially in education, because ethics and morality are closely intertwined. An ethical framework must be based on the ethical and moral values of a given society or country, rather than imposing a global framework for all countries. That said, this does not preclude adopting minimal ethical principles at a global level.
AI is now included as a competency for the professional development of teachers, but again, there is no clear definition of how this should be implemented. I understand that one could argue it is up to national and supranational educational authorities to define teacher training plans and decide how to incorporate AI competencies. However, it would have been beneficial for UNESCO, as a global consultative body, to urge national governments and supranational institutions to adequately fund development and training plans for teachers in public education.
Once again, the teacher and student frameworks seem like a kind of 'wishful thinking'—well-intentioned but potentially ineffective if sufficient funds are not allocated for training, especially for teachers who are the ones responsible for helping students develop their AI competencies.
References:
Ng, D T K, Lok Leung, J L, Wah Chu, SK, Shen Qiao, M (2020), Conceptualizing AI literacy: An exploratory review, Computers and Education: Artificial Intelligence, Volume 2, 100041, ISSN 2666-920X. Available at: https://doi.org/10.1016/j.caeai.2021.100041 (Accessed 20th March 2024).