Beyond automated judgement: reframing  Intelligent Computer-Assisted Language Assessment (ICALA) as an assessment ecology

The rapid integration of artificial intelligence into language assessment has accelerated the adoption of Intelligent Computer-Assisted Language Assessment (ICALA) across educational contexts. Existing research has largely focused on technical performance, efficiency, and scoring accuracy, often treating AI as a neutral enhancement to assessment practice. This Special Issue editorial argues that such framings are insufficient. Drawing on seventeen empirical studies, the editorial reframes ICALA as an assessment ecology, a socio-technical configuration in which technological affordances, learner affect and agency, teacher mediation, institutional norms, and equity considerations are dynamically intertwined. Through thematic synthesis, the editorial demonstrates that ICALA systematically reshapes the emotional conditions of assessment, learner participation and autonomy, teacher professional well-being, and experiences of fairness and integrity. The analysis highlights that affective benefits, agency, and engagement are contingent on assessment design, transparency, and mediation, rather than inherent properties of AI systems. Methodological patterns across the Special Issue are examined to identify limitations in current approaches and to outline directions for future research capable of capturing relational, longitudinal, and ethical dimensions of AI-mediated assessment. The editorial concludes by setting an agenda for ICALA research that moves beyond automation and efficiency towards principled questions of validity, equity, and responsibility in increasingly algorithmic assessment environments.