UNESCO has published a document setting out the competency framework for teachers on artificial intelligence (AI), which shows what educators need to know, value and be able to do in order to use this technology with judgment. The document, the first of global scope, proposes fifteen competencies organized into five areas and three levels of training. Its aim is for teachers to be able to decide when, how and for what purpose to incorporate AI, keeping the human relationship at the center of learning.
Artificial intelligence is already in schools, where it drafts texts, summarizes readings, corrects exercises and suggests decisions based on large volumes of data. Its presence opens up new possibilities, but it also forces a rethink of how teaching is done and what role teachers should play. In this regard, UNESCO's framework, known as AI CFT, offers a reference to guide that preparation.
AI offers opportunities to improve teaching, personalize learning and lighten administrative tasks. However, it can also weaken people's autonomy, violate privacy, deepen inequalities and reduce education to what a machine can process.
The OECD identifies 200 AI use cases for governments
Teacher preparation is therefore decisive. Even so, according to the survey cited in the document, in 2022 only seven countries had AI training frameworks or programs for educators. Many teachers face this technology without guidance that would allow them to assess its possibilities, recognize its limits and decide when it is appropriate to use it.
Unlike other digital technologies, AI can emulate human behavior. That characteristic poses four challenges that directly affect teachers' work. The first is its capacity to take on tasks and decisions that previously belonged to people. Excessive dependence can erode the educator's own competencies and reduce their professional autonomy.
The second relates to data. Training systems can involve the use of personal information, often without effective consent from those who provide it. The third is the probabilistic nature of its outputs, since the same input can generate different responses. Each result therefore requires critical evaluation, even when it seems convincing or is well written.
Finally, AI's capacity for adaptation makes it a factor of social transformation. Students need to learn to understand these changes and their consequences, as well as to become familiar with the tools.
The AI CFT is based on a human-centered vision. AI must strengthen people's capacities, be equitably accessible, be explainable and safe, and remain under human control and responsibility.
This approach implies that the technology must not replace teachers or substitute for the full range of their functions. With the incorporation of AI, the pedagogical relationship comes to include three elements: teacher, student and artificial intelligence, which requires protecting the rights, responsibilities and role of educators.
Secondly, the tools must be validated before teachers are required to apply ethical principles in their use. Teachers need adequate conditions to work responsibly, without bearing governance obligations that exceed their role. The framework combines five areas of competency: a human-centered way of thinking, AI ethics, the foundations and applications of AI, AI pedagogy and the use of AI for professional development. Each area is developed across three progression levels: acquire, deepen and create, which make up a total of fifteen competencies.
The acquire level corresponds to the basic literacy that all teachers need. It involves understanding that AI is directed by people, knowing its ethical risks, understanding how models are trained and learning to make the most of its pedagogical benefits.
Deepen means integrating AI into educational practice responsibly. At this level, teachers must comply with local regulations, protect privacy and intellectual property, and expand their knowledge through shared learning with other colleagues.
Create, in turn, is aimed at expert teachers, capable of participating in the design of tools, contributing to the development of ethical rules and innovating in teaching.
This progression is not necessarily linear or uniform. A teacher may have advanced knowledge of the technical foundations of AI and still be at a basic level on ethical issues. The framework therefore recommends assessing competencies in each area and using those diagnoses to personalize training.
Defining competencies is only the first step. To bring them into the classroom, the document recommends that each country analyze the costs and benefits before promoting the adoption of AI. This technology is unlikely to solve on its own problems such as shortcomings in school infrastructure or the shortage of teachers.
UNESCO also proposes offering well-funded training, recognizing innovative practices and guaranteeing affordable access to previously validated tools. This is in addition to the need to adopt frameworks adapted to the local context, preferably under the leadership of governments.
As references, the document cites the strategy of the Republic of Korea, the European AI4T project and Singapore's platform for educators. These experiences illustrate different ways of supporting teachers in incorporating AI into their work.