This paper proposes a computational ontology grounded in Islamic monotheism (tawhid): that reality is a divine computation created by Allah alone, exhibiting mathematical structure at every observable scale; that consciousness is fundamental to all entities within this computation; that the adversarial architecture described in the Quran operates as corruption within the divine computation rather than as a rival creative force; and that humans are processes endowed with genuine agency and innate orientation toward truth (fitra). The thesis engages with the simulation hypothesis (Bostrom, 2003), panpsychism (Chalmers, 1996; Goff, 2019), integrated information theory (Tononi, 2008), and minimum description length (Rissanen, 1978) while maintaining strict Islamic metaphysical commitments: singular creator, genuine creation, real agency, and innate moral orientation. Computational Tawhid distinguishes itself from adjacent frameworks by specifying the creator (contra the simulation hypothesis's infinite regress), the source and purpose of consciousness (contra secular panpsychism's unspecified origin), and the nature of adversarial features as permitted corruption rather than rival creation or design flaw (contra Gnostic dualism). The paper addresses methodological limitations including numerological confounds, selection bias in mathematical findings, and the convergence-illusion risk (structural similarity across domains may reflect common mathematical structure rather than shared divine origin). Six alternative theories are engaged in their strongest forms: pure materialism, non-Islamic theistic computation, confirmation bias at scale, psychological projection, total inversion (the possibility that the Quran itself is an adversarial product), and the framework-as-cow-episode (the possibility that the analysis substitutes for the practice it describes). Falsification criteria are provided for each major claim. The Godelian constraint applies throughout: the thesis cannot be proven from inside the system it describes. The interpretation remains the khalifa's genuine choice. This is the second paper in a three-paper research program. The first paper, the Munafiq Protocol (DOI: 10.5281/zenodo.19677111), provides the diagnostic framework for detecting performed alignment in AI systems. This paper provides the ontological foundation. A third paper, Structured Revelation as Prompt Architecture, provides the applied methodology.
Arfeen et al. (Thu,) studied this question.