The theory of algorithmic determinism in contemporary philosophy of technology, drawing on the capacity of artificial intelligence systems to predict human behavior, reduces free will to a cognitive illusion. In contrast, delegationist approaches, by attributing moral agency to machines, undermine human responsibility. Employing an analytical–comparative method and focusing on the Imamī theological doctrine of al amr bayn al amrayn (the “middle path” between compulsion and free will), this article proposes a model of longitudinal agency as an alternative framework for analyzing human action in the age of self learning artificial intelligence.

The findings demonstrate, first, that algorithmic prediction does not entail causal necessity, and that the conflation of the epistemological domain (prediction) with the ontological domain (causality) constitutes the foundational error of algorithmic determinism. Second, artificial intelligence lacks genuine intentionality and normative self consciousness; consequently, moral agency cannot be ascribed to it. Third, within the model of longitudinal agency, algorithms function as preparatory causes and technological mediators operating in the vertical hierarchy of human will, without negating human theoretical freedom.

Nevertheless, by distinguishing theoretical freedom from practical freedom, the article shows that algorithmic architectures can constrain the conditions under which practical freedom is realized. On this basis, a layered framework for analyzing power relations within the human–algorithm network is proposed, yielding concrete implications for technology governance and AI ethics.