<![CDATA[The problem of determinism in machine learning is becoming increasingly relevant as predictive analytics, recommendation systems, and algorithmic profiling model human behavior as predictable statistical patterns. Modern studies of the philosophy of technology generally address this issue within the framework of secular rationalism. At the same time, the contribution of classical Islamic theology to the critique of digital determinism remains limited. This study aims to analyze the problem of algorithmic determinism by reconstructing the Ashʻarite and Muʻtazilite concepts of free will and to assess their relevance as a critical framework for the metaphysical assumptions of machine learning. This research is a literature study with a philosophical-theological approach, grounded in a hermeneutic reading of primary kalām texts and a conceptual-critical analysis of contemporary machine learning literature. The results show that determinism in machine learning is both operational and epistemic, as algorithmic predictions rely on probabilistic correlations rather than final causality in human actions. The Ashʻarite perspective, through the concept of kasb, shows limited compatibility with digital predictive systems by distinguishing between the creation of actions and the subject’s moral responsibility. In contrast, the Muʻtazilite affirmed human rational autonomy and the causal efficacy of actions, thereby demonstrating the limitations of algorithmic deterministic claims in explaining ethical freedom. A comparison of the two suggests that machine learning does not eliminate free will but only enhances the capacity to predict behavior. This research confirms that classical kalām can serve as a critical framework for evaluating the metaphysical assumptions of contemporary digital technology.]]>
