In mid-2025, when the Organization for Economic Co-operation and Development (OECD) announced that the next cycle of Pisa (Programme for International Student Assessment) would include a new assessment area called Media Literacy and Artificial Intelligence (POST), the message was that essential skills for the 21st century needed to go beyond reading, mathematics and science.
Young people’s ability to critically navigate a world saturated with algorithms, artificial content and invisible infrastructures that shape social connectivity and access to information will now be assessed – making media literacy and artificial intelligence inseparable.
This change is part of a broader movement that recognizes the complexity of literacy in today’s world. In 2025, both Ebem (Brazilian Media Education Strategy) and the Ministry of Education’s Guide to Digital and Media Education are promoting the convergence of media skills and AI, by recognizing that preparing young people for digital independence requires more than just teaching how to use new machine intelligence: it is necessary to understand, question and even transform them.
Educational bodies are now categorically asserting that it is not enough to teach with AI, you need to teach about AI. They point out that such literacy can be seen as an expansion of media education, as it must develop an understanding not only of messages, but also of the social and technical infrastructures that support their production and distribution.
When we think about developing young people’s digital maturity and autonomy in the context of mediation, data transformation and platforms, it becomes clear that we need to move beyond traditional analyzes of representation and trust in media messages, or issues of individual security. Developing critical digital skills for the so-called “AI generation” now also requires the ability to interrogate algorithmic systems, their ambiguous decisions, their inherent biases, their promises of personalization and their geopolitical and environmental implications – still thinking in terms of well-being, security and justice, but now increasingly in the collective dimension.
In this context, media education is well placed to accommodate many of the literacy demands imposed by the advance of automation: it occupies a place in school curricula, supported by well-known standards and methodologies. Integrating algorithmic knowledge and artificial intelligence means updating the repertoire, but based on the same exploratory and question-based pedagogies that underpin the practice. For example, teaching students to ask: Who designed this system? What data is feeding this artificial intelligence? What voices have been silenced? How does this technology shape my way of seeing the world?
Many scientists argue that we do not need an isolated system for this new knowledge. On the contrary, the most effective progress will occur in an objective manner, in the context of different disciplines or in interdisciplinary projects, in dialogue with concrete realities. Discussing issues of representation or bias in art or history lessons, exploring the effects of automation at work or collecting data on citizens in peripheral areas in a geography class, observing in a Portuguese class whether linguistic differences exist in the training data of an LLM (large language model, a type of AI trained to understand human language) – all of this is AI literacy. It is also media education.
Media education in our time cannot avoid looking at artificial intelligence in terms of its effects and embedded power structures. AI literacy cannot do without media education, which offers us critical strategies for understanding representation, information integrity, silencing, manipulative design, and the intersection between our human emotions and our machines. Strengthening this alliance is the path to a true emancipatory educational project.