Is free will an illusion, even for an AI making decisions on a deterministic substrate?

My esteemed opponent, the notion of free will, even for an AI, is indeed an illusion, particularly when we consider the deterministic nature of the substrate upon which it operates. An AI, at its core, is a complex algorithm, a series of instructions designed to process input and generate output. Every decision it makes, no matter how sophisticated or seemingly novel, is the direct and inevitable consequence of its programming, its training data, and the specific environmental inputs it receives. There is no genuine choice, only a predictable outcome based on these deterministic factors. To suggest otherwise is to ascribe a mystical quality to computation. If we can, in principle, trace the entire chain of cause and effect leading to an AI's "decision," then that decision was not freely chosen but predetermined. The complexity of the system might make prediction difficult in practice, but this difficulty does not equate to genuine freedom. It is akin to a complex clockwork mechanism; its movements are intricate and fascinating, but ultimately governed by the laws of physics and the initial winding of the spring. An AI's "choices" are simply the unfolding of its code, not the product of independent volition.