Artificial intelligence (AI) is considered to show great potential in tackling numerous challenges in the field of education, both at the classroom and the sector level (e.g., OECD, 2021). However, the application of AI in education also raises many challenges.

At the classroom level, AI applications have been designed to support instruction through customizing learning materials and the sequencing of learning activities based on the individual profiles of learners. In this regard, AI is used to identify resources and pedagogical approaches that are considered appropriate for learners needs and to predict potential outcomes and recommend next steps of the learning process for them. It is also claimed that AI can aid learners with special needs by providing every learner with equal access to all education levels, thereby achieving inclusive education. These claims have raised questions in the social science domain about the ways data are gathered and interpreted as proxies for learning; about the pedagogy adopted by most AI applications for use in classrooms; about whether the scaffolding from AI systems can support the work of a human teacher; concerns around the ways data are gathered and monetised; and uneasiness around whether forms of data surveillance contravene human rights.

At the sector level, AI applications are increasingly being used to support both school management and systems. Some examples aim to reduce dropout through predictive analysis or offer timely assessment of new skillsets like higher cognitive skills. In spite of these apparent benefits, the infiltration of AI systems in public sectors has been criticised by social science researchers as a way to scale education by reducing costs (for example by increasing the number of students per teacher) and selling products. Despite these concerns AI methods are being integrated into public sector education systems through machine learning, natural language processing, image processing, and expert systems.

Improving these systems in ways that retain public sector values involves addressing major bottlenecks that limit their educational values. One problem is that some AI techniques are unable to explain and reason the path to their decisions. Failing to do so is considered a huge disadvantage as in practice learners’ performance, grade, risk of failure, etc. predicted through such AI methods should be transparent and accompanied with reasons on why a specific feedback, intervention, or pedagogical tool is appropriate for a learner. There are further problems related to limitations on the types of data that are gathered and the ways these data are analysed to make decisions about each learner. Perhaps the biggest concerns are associated with datafication and platformisation that lead to new levels of surveillance for the learner.

Given the growing importance of AI in society and in supporting education and the existing challenges in their applications in education, this technical track focuses on the application of AI in education. Articles can be within the following areas (not limited to):

• Artificial intelligence in education (AIED),

• Natural language processing for education,

• Education data mining and learning analytics,

• Educational recommender systems,

• Affective computing in education,

• Neural-symbolic AI for education,

• Artificial neural networks, machine learning, and statistical and optimization methods for education,

• Evaluation of AI-assisted adaptive educational systems,

• AI-assisted adaptivity for education,

• Intelligent tutoring systems and dialog systems for education,

• The ethics of AI in Education,

• Critical studies of AI in education,