1467-9744 Zygon: Journal of Religion and Science 1467-9744 Open Library of Humanities 10.16995/zygon.24448 Article Ibn Sīnā’s Philosophy as Framework for Understanding Artificial Intelligence: Ontology, Epistemology, and Machine Cognition’s Boundaries Shahab Hasan hasan.shahab@cybertrend-intra.com 1 https://orcid.org/0000-0002-3407-760X Gama Cipta Bakti ciptabaktigama@sadra.ac.id 2 Chairman, Indonesian Data Science and AI Association (ASDASI), Jakarta, Indonesia Assistant Professor, Quran and Islamic Philosophy, Sekolah Tinggi Agama Islam Sadra, Jakarta, Indonesia 12 10 2026 2026 61 3 815 833 Copyright: © 2026 The Author(s) 2026 This is an open-access article distributed under the terms of the Creative Commons Attribution 4.0 International License (CC-BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. See http://creativecommons.org/licenses/by/4.0/. Articles published by the Open Library of Humanities may also contain other CC-licensed material, or material used or reproduced under an exception or limitation to copyright. Such material is not subject to the article’s license. It is either marked with its own license, or marked that it has been reproduced with permission or under fair-use.

As artificial intelligence (AI) systems become increasingly sophisticated, fundamental questions arise: Do they possess genuine existence? Can machines truly think? This article addresses these questions through Ibn Sīnā’s (Avicenna, 980–1037) philosophical framework. Drawing on his distinctions between essence and existence, substance and accident, and natural (tabīʿī) versus artificial (ṣināʿī) beings, it argues that AI is fundamentally an artificial entity whose existence remains accidental, derivative, and externally dependent. Epistemologically, Ibn Sīnā’s theory of intellect encompasses abstraction (tajrīd), intuition (al-hads), and illumination by the Active Intellect. Human intelligence, arising from the rational soul (nafs), enables understanding, self-awareness, and participation in intelligible realities whereas AI operates as a closed cognitive system incapable of genuine understanding, consciousness, or universal knowledge. While rejecting strong AI and machine personhood as metaphysically untenable, the article advances a constructive view: precisely because AI lacks consciousness and moral agency, it can relieve human cognition from instrumental tasks, enabling cultivation of theoretical intellect, practical wisdom (ḥikmah), and ethical responsibility under human governance.

 Introduction

The rapid progress in artificial intelligence (AI) compels us to ask not just how it works but what it truly is. This ontological inquiry seeks to understand AI’s fundamental nature, moving beyond a simple description of its functions. The core philosophical questions remain: Can machines genuinely think? What and how is human intelligence different from AI? Can AI truly replicate human intelligence? And could AI become conscious someday? What standards can we use to judge the possibility of AI personhood and moral responsibility? The answers to these questions will shape how we think about AI ethics and regulation. This is not just a technical puzzle; it is a deeply philosophical one, because the very definitions of “intelligence” and “thought” are fraught with scientific ambiguity and carry profound moral, religious, and legal weight.

Even a seemingly straightforward definition of AI, like the one Google offers—a field of science dedicated to building machines capable of reasoning, learning, and acting in ways that typically demand human intelligence or handling data beyond human analytical capacity (Google Cloud, n.d.) —begs the question of what constitutes “human intelligence” in the first place.

More generally, AI is often described as a branch of computer science aiming to imbue computers with intelligence to assist humans (Cycleback 2018, 9). Modern AI systems typically incorporate a knowledge base, a network of interconnected facts, and an inference engine, which allows the system to draw conclusions based on data (Haenlein and Kaplan 2019). Some researchers define AI as “the ability of digital computers or computer-controlled robots to solve problems that typically require human beings’ higher cognitive processing abilities” (Copeland 2026).

Looking at these various perspectives, it is clear that AI is both a science and a technology. While its scientific foundations have been in place for some time, the field of AI formally emerged in 1956 at the Dartmouth Summer Research Project on Artificial Intelligence, organized by John McCarthy and Marvin Minsky. Its technological development then accelerated alongside the advancement of computers as data processors operating on defined models (Haenlein and Kaplan 2019).

Fundamentally, AI rests on mathematical concepts like computation and algorithms, evolving into a technology driven by three key elements: processing power and infrastructure, programming paradigms, and, crucially, data (Alawijeh and Al-Mesilini 2025). The “cognitive” aspect, the defining characteristic of AI, is inextricably linked to the technology itself. Therefore, understanding AI as a technology necessitates examining the underlying and emerging scientific principles that shape it.

The debate surrounding AI often boils down to two opposing definitions: “weak AI,” which acknowledges the potential for machines to act intelligently, and “strong AI,” which posits that these actions constitute genuine intelligence, even suggesting that certain artificial computations are thinking. This distinction highlights the deeply philosophical nature of defining intelligence in the context of AI, moving beyond purely technical considerations. The very existence of this debate reveals that operational definitions of AI focusing solely on its capabilities are insufficient to address the fundamental question of its nature. If strong AI proves to be correct, it would challenge anthropocentric views of intelligence and consciousness, with far-reaching moral, religious, and legal implications (Shareef 2023, 5–6).

 Why Ibn Sīnā Matters in the Age of AI

Recent scholarship has increasingly called for the inclusion of non-Western philosophical traditions in contemporary debates on AI. For instance, Biliana Popova’s (2020) article in Zygon: Journal of Religion and Science suggests that Islamic thought should inform discussions about AI development. She argues that more scholars from the Islamic philosophical tradition should participate in debates about AI development and its implications, particularly since many Muslim countries are leaders in AI development and its application in daily life. Islamic philosophy, with its diverse schools and related ontological, epistemological, and ethical beliefs, offers valuable insights for discussing AI development. However, Popova’s account does not engage with the philosophy of Ibn Sīnā, whose systematic and far-reaching contributions remain underexplored in this context. Yet, as Peter Adamson observes, Ibn Sīnā is not just a compiler of Greek wisdom. He is a thinker of astonishing originality who reshapes everything he touches—from logic to science, from psychology to metaphysics. To understand philosophy in the Islamic world after 1100, you cannot avoid Ibn Sīnā. He is the most famous and influential philosopher of the Islamic world, whose thought came to define the philosophical tradition for both later Islamic thinkers and Latin Christian philosophers like Thomas Aquinas (Adamson 2016, 636–41; McGinnis 2010, 9–16).

This article explores questions surrounding AI through the philosophical lens of Abū ʿAlī al-Ḥusayn ibn Sīnā (980–1037), known as Avicenna. He stands as a key figure in the Greco-Arabic philosophical tradition, offering a comprehensive view of logic, physics, cognition, mathematics, psychology, and metaphysics, detailed in works like the Kitāb al-Shifāʾ (The Book of Healing) (Ibn Sīnā 1997) and Al-Ishārāt wa al-Tanbīhāt (Remarks and Admonitions) (Ibn Sīnā 2002).

But why turn to Ibn Sīnā when thinking about AI? There are at least two compelling reasons. First, given AI’s interdisciplinary nature, Ibn Sīnā’s thought provides a relevant framework for analyzing AI intelligence. Ibn Sīnā possessed extensive knowledge in philosophy, psychology, science, cognition, mathematics, and anatomy, especially brain anatomy (McGinnis 2010). Second, his comprehensive insights offer a strong foundation for examining the nature of both human and artificial intelligence. Ibn Sīnā long ago discussed the differences between natural (ṭabī‘ī) and artificial (ṣinā‘ī) entities.

Recent popular and academic discussions have begun to recognize this relevance. Abigail Tulenko (2024) wrote in Scientific American under the title, “What Philosopher Ibn Sīnā Can Teach Us about AI.” Tulenko suggests exploring the thoughts of Ibn Sīnā to address some of the philosophical problems of AI. Ibn Sīnā’s thinking presents a thorough framework for defining what it means to think and be conscious, a central question in current debates about the nature of AI. Ibn Sīnā developed an intricate philosophical system addressing reality, the soul, and knowledge. His philosophy regarding the soul, intellect, and how we gain knowledge proves particularly relevant. He delves into the essence of intelligence, consciousness, and cognitive processes (Norman 2021).

Beyond his detailed philosophical framework for science and technology, Ibn Sīnā considered the distinction between the natural and the artificial. Contemporary studies acknowledge Ibn Sīnā’s relevance when comparing natural intelligence and AI (Alawijeh and Al-Mesilini 2025). He also discussed the subjects of various sciences, raising questions about the mode of existence for each science’s subject matter. This includes discussions related to mathematics, which encompasses algorithms and computations—the core of AI intelligence (Ibn Sīnā 1960, 18–19; Marmura 2005, 249–56; Zarepour 2019, 95–126).

Nevertheless, the engagement of classical philosophy with AI is not unprecedented. Aristotelian frameworks have been employed to analyze AI ontology through hylomorphism and the natural–artificial distinction (Dreyfus 1992; Coeckelbergh 2020), while Kantian approaches have addressed questions of autonomy and moral agency (Vallor 2016). Phenomenological critiques, particularly Martin Heidegger’s concept of Dasein and embodied cognition, have demonstrated AI’s inability to achieve genuine understanding (Dreyfus 1992; Wheeler 2005). These approaches have generated valuable insights, yet they often remain confined to specific dimensions of the problem.

In contrast, this article argues that Ibn Sīnā’s philosophy offers three distinctive contributions. First, it provides a systematically integrated framework where ontology, epistemology, and ethics are interconnected: AI’s ontological status as ṣinā’ī (artificial) necessarily entails its epistemological limitations and ethical implications. Second, Ibn Sīnā’s detailed faculty psychology—specifying five internal faculties with precise cognitive functions—allows us to map which aspects of cognition AI can simulate (data processing, pattern recognition) and which it categorically cannot (estimation of non-sensory meanings, abstraction to universals). Third, his theory of the Active Intellect (‘aql al-fa‘āl) provides a unique metaphysical explanation for why AI cannot achieve universal knowledge: it has no access to the immaterial source of intelligibles.

Furthermore, unlike approaches that remain primarily critical, Ibn Sīnā’s framework allows for a constructive reorientation of the human–AI relationship. Precisely because AI does not possess independent agency or intellect, it can be understood as an instrument that supports human cognitive development. In this sense, AI may function as a means of cognitive extension, enabling human beings to redirect their intellectual capacities toward higher forms of theoretical knowledge as wisdom (ḥikmah) and ethical action.

While Aristotle provides a foundational hylomorphic framework for understanding form and matter, Ibn Sīnā extends this metaphysical structure by introducing a more rigorous distinction between essence (māhiyyah) and existence (wujūd), thereby allowing a more precise ontological classification of artificial entities such as AI. Furthermore, while Kantian philosophy locates the limits of AI in the absence of autonomy and moral agency, Ibn Sīnā’s framework deepens this critique by demonstrating that such limitations originate at a more fundamental level—namely, in the absence of a metaphysical ground for intellect, self-awareness, and universal knowledge. This article thus demonstrates how a rigorously developed premodern Islamic philosophical system can address contemporary technological questions with conceptual resources that neither Aristotelian, Kantian, nor phenomenological frameworks fully provide.

 Philosophical Framework of Ibn Sīnā and Contemporary Artificial Intelligence Systems

Ibn Sīnā’s detailed theory of existence, with its careful distinctions between what something is and the fact that it is, between what must exist and what could exist, and between physical and non-physical things, offers a strong philosophical foundation for considering the nature of AI.

Our investigation begins by laying out Ibn Sīnā’s basic categories of being. Then, we consider how modern AI systems might fit within this framework. The core question is whether AI can achieve the kind of real existence Ibn Sīnā ascribes to true intellectual beings, or whether AI is permanently limited to being a modification of existing things.

The Difference between Essence and Existence

A key point in Ibn Sīnā’s philosophy is the difference between essence (whatness/māhiyyah) and existence (being/wujūd). For Ibn Sīnā, essence is what defines a thing—its fundamental nature. Existence, on the other hand, is the fact that the thing actually is (Ibn Sīnā 1960, 42; Belo 2009).

Ibn Sīnā believes that essence, in itself, is neither existent nor non-existent. The essence of “horse,” for example, stays the same whether we are talking about a real horse, a horse in our imagination, or just the idea of a horse. Existence, therefore, is something added to the essence of things that do not have to exist (Sīnā 2002, 47; Marmura 2005, xxii–xxiii).

This idea is important when we think about AI. Do AI systems truly exist in a way that is distinct from their underlying structure—their code, architecture, and data? Or are they just complex arrangements of essence without their own independent being?

 Substance and Accidence

Ibn Sīnā’s theory of categories draws a key distinction between substance (jawhar) and accidence (‘araḍ). A substance exists independently and acts as the subject for accidents, while accidents are modifications or properties that exist within substances (Ibn Sīnā 1960, 71; Belo 2009). This distinction is crucial for considering whether AI systems can achieve genuine, independent existence or remain simply accidental modifications of their material components.

Beyond the technological aspects of AI hardware, it is important to examine the knowledge derived from human mental processing, where mathematics (computation and algorithms) form the foundation of AI. Ibn Sīnā categorizes theoretical knowledge into natural science (ṭabī‘iyyāt), mathematics (riyādhiat), and metaphysics (Ilāhiyyāt) (Ibn Sīnā 1960, 11; Marmura 2005, xix–xx).

For him, mathematics is a distinct field with its own specific subject matter: quantity (kam). He defines quantity as something “abstracted from matter or from anything that has quantity” (Ibn Sīnā 1960, 4). This abstraction is significant because it directly separates mathematics from physics, which deals with objects in motion and at rest. Ibn Sīnā argues that quantities, both continuous and discrete, are fundamentally accidents, not independent substances (Zarepour 2019, 55). This means they need a substratum, a material object, to exist in the external world. For instance, “two” exists in “two apples” or “two books,” not as an independent entity. In Ibn Sīnā’s view, quantity is conceptualized (abstract concept) as an accident of matter; its existence depends on material substances. The process of “mental separation” (abstraction) is at the heart of how humans study these objects. Although mathematical objects are mixed with specific types of matter in the external world (e.g., an apple circle), the human mind can separate them from these particulars. One can understand a circle without it being specifically made of apple or ball (Ibn Sīnā 1960, 118; Marmura 2005, 84–87; Zarepour 2019, 71).

A tension arises here: How can an accidental property tied to a specific material instance (e.g., the “two” of “two apples”) be a universal object of abstract thought? This highlights the vital role of the human mind’s capacity for abstraction in Ibn Sīnā’s philosophy. Mathematical objects have no autonomous extramental nonmaterial existence (Marmura 2005, xix). Numbers are not immaterial substances themselves, but human intellect makes them so through a process of mental detachment from their particular material embodiment, making universal mathematical study possible (Zarepour 2019, 141–52).

According to Ibn Sīnā’s framework, mathematical objects, which are at the heart of AI algorithms, are fundamentally accidents (‘arad). When we consider the origin and purpose of AI, and how its “intelligence” is interpreted, the role of the human mind becomes fundamental. The algorithms, data structures, and architecture of AI are the result of the abstract thinking, reasoning, and creativity of the human mind. The mathematical concepts that underlie AI, such as logic, number theory, and probability, are products of human intellectual activity (Ibn Sīnā 1960, 18–19; Zarepour 2019, 152–58). AI’s origins lie in human intellect, where mathematical and logical concepts take shape. The very essence of AI algorithms exists first as abstract ideas within the human mind.

When a scientist designs algorithmic architecture, the concept is present in his or her mind as an intellectual disposition. It is an accident inherent in the substance of the human rational soul. When the algorithmic code is executed on a silicon chip, its ontological status changes completely. It ceases to be “knowledge” and becomes a particular matter arrangement. Based on Ibn Sīnā’s principle that mathematical objects (such as the binary numbers 0 and 1) cannot be separated from materiality, the algorithm and data here become accidents that are completely dependent on hardware. AI, in its functional form, is an accident of the hardware that runs it. The functional existence of AI technology depends on its hardware. Without silicon and other physical components, algorithms and data cannot operate or manifest. Therefore, the hardware substance, like silicon, provides the material support for AI’s functional existence.

 Natural versus Artificial Entity

Even though Ibn Sīnā lived long before we started thinking about AI, his ideas provide a basis for telling the difference between things that have their own internal drive (like nature or a soul) and things that are made by external forces. Ibn Sīnā draws a critical line between natural entity (ṭabī‘ī) and artificial entity (ṣinā‘ī). Natural entities have their own inherent nature (ṭabī‘ah) that drives their motion, development, and how they become what they are. They exist according to their core characteristics and achieve their potential through these built-in drives (Ibn Sīnā 1997, 158–59; Alawijeh and Al-Mesilini 2025).

In his philosophy, Ibn Sīnā focuses quite a bit on natural entity (ṭabī‘ī). He defines “nature” (ṭabī‘ah) as more than just the physical environment. It is the main reason (mabda’ awwal) natural things move and rest. This reason is part of what the thing is, not just something that happens to it from the outside. He clearly separates motion that comes from a thing’s own internal drivers (like its nature or soul) from motion caused by outside forces or “violence” (think of a thrown stone versus a stone that falls on its own) (Ibn Sīnā 1983, 31; Arif 2007).

In the natural order, an object does not automatically have the ability to transform into any other object. Ibn Sīnā suggests that the potential to “become” implies an internal state within an object, propelling it toward realization and perfection. However, the composition of matter differs between artifacts and natural entities (Ibn Sīnā 1983, 34–37). For instance, a teak tree seedling possesses the inherent capacity to mature into a large tree, yet it lacks the innate potential to transform into a chair. However, through human intervention, the seedling undergoes a series of transformations to become a chair. In essence, wood does not inherently strive to become a chair, and the chair’s form is not the wood’s objective; rather, it is the carpenter’s objective (Mohaddes 2020).

In human-made objects (artifacts), form simply refers to physical shape or structure. When humans create something, they arrange pieces of matter and decide how these parts relate to each other. Many modern philosophers argue that nature works the same way—that natural things also have forms that are just arrangements of matter, similar to human artifacts. However, Ibn Sīnā believes that “form” in nature is something much deeper and more significant than mere physical shape. According to Ibn Sīnā, form in natural things represents a higher reality. The physical shape and organization we observe are actually just outward expressions or manifestations of this deeper reality. “The natural world is capable of a synthesis that leads to transformation, whereas the artificial world cannot achieve this, but rather, it can only aggregate” (Ibn Sīnā 1997, 158–59).

In artificial (human-made) things, what we consider the “essence” depends on the form, not the material. This is why artificial objects do not have a true essence—their composition is based on human convention rather than natural reality (Ibn Sīnā 1983, 34–37). Consider a wooden chair made of several pieces of wood assembled into four legs, a seat, and a backrest. If we take apart this wooden chair and rebuild the exact same structure using iron instead, we still call it a chair. The material changed completely, but because the form (shape and structure) remained the same, we recognize it as a chair. This shows that we identify the chair by its form, not its material.

Now, let us reverse the scenario. If we keep the same wood (the material) but change the form—for example, by reshaping that same wood into a blackboard—it is no longer a chair. The essence has completely changed, even though the material is identical. In human-made objects, the creator (the “agent”) is responsible for creating the form, not the material. For example, a carpenter creates the form of the chair but does not create the wood itself (the material). Wood exists independently of the carpenter’s work. The connection between a chair’s shape and its purpose is a connection only made up by the minds of the maker. This connection is true from the perspective of the maker who created the shape for this use, not from the perspective of the shape (form) itself or what it is really used for (Mohaddes 2020; McGinnis 2025).

Artificial entities exist because of outside forces and human skill. They do not have their own internal drives for motion and change. Instead, they depend on external agents for their shape and function and how they keep working. While artificial entities are real, they are a different kind of being—one marked by dependence, being a tool, and external control (Marmura 2005, 218, 295).

Based on these parameters, an artificial (ṣinā‘ī) object is defined as the result of a volitional act (derived from the choice of an external agent), yet uniform in its operation. A clepsydra (water clock), for example, is an artificial entity (ṣinā‘ī). Its creation involves human will and intelligence, but once the clock is completed and running, it operates automatically and uniformly according to its mechanical design. It has no freedom to change its own rhythm (McGinnis 2025).

This difference is very important for making sense of AI systems. They clearly fit into the artificial entity category, created by human programmers and lacking their own internal ways to grow or become what they could be. Algorithm design highlights AI’s dependence on human creators. Even when AI generates its own algorithms, these changes arise from externally induced variations (e.g., adjustments in model parameters, network weights, or new algorithm structures) driven by data input, the computing environment, or human-programmed objectives. This “accidental motion” is governed by external factors (humans, electricity, code) rather than “substantial motion” originating from within the AI itself. And AI automation is the result of actions that originate from the choices of external agents but are uniform in their operation.

 Key Differences between Natural Human Intelligence and Artificial Intelligence

The approach taken by Roger Penrose is useful to begin this discussion. According to Penrose (2002, 28), “I would say that in ordinary usage the word ‘intelligence’ requires ‘understanding,’ and I would say that in the normal usage ‘understanding’ requires some ‘awareness.’ So, I would say that for an entity to be ‘intelligent,’ in the ordinary usage of the word, it would have to be aware, it would have to be conscious.” This tri-conceptual framework resonates remarkably with Ibn Sīnā’s ontological distinction between natural and artificial entities, developed nearly a millennium earlier.

1. Self-awareness and self-discovery

Ibn Sīnā argues that natural intelligence, at its highest level (human level), is self-aware. Ibn Sīnā draws a parallel between the human body (material) and the soul (nonmaterial/spiritual) to elucidate self-awareness. He likens this relationship to that of a body and its clothing. The difference, however, is that clothes can be removed when not needed, whereas the body cannot. In al-Shifā’, Ibn Sīnā (1997, 324–25) writes:

We are not accustomed to detaching the body from ourselves. Even if one does not consider the entire body as the self, they still believe that certain organs—such as the head, hands, or feet—are part of the self. If one were to consider the entire body or its organs as the self, this feeling should be constant. However, in reality, humans do not always perceive their body or its organs as themselves. For example, we do not feel the presence of the heart or brain. We learn about these organs through study, information, or direct observation during surgery. We do not inherently feel that these organs are the self.

The body and its organs are not essential, as we do not always perceive them as part of ourselves. Therefore, the “I” is not the body. The “I” is something whose presence must be constantly felt because it is essential. It is the “I” that moves, walks, and acquires knowledge. The unified and indivisible self is the “I.” If one is convinced that there is an ‘I’ within them, then what is this “I”? Is it the soul? “I” is the soul because it aligns with the attributes of the soul. Even if we do not know the precise term, we understand its essence. The substance with these characteristics has been proven to align with the definition of the soul.

To illustrate that the human soul is a spiritual (nonmaterial) substance, Ibn Sīnā proposes the “flying man” thought experiment. Imagine a person created instantaneously, without any interaction with the external environment. This person would lack experience, memories, and imagination, as these depend on the physical world and sensory input. Despite this deprivation, Ibn Sīnā contends that the “flying man” would still be able to affirm their own existence, demonstrating consciousness. This suggests that the soul, which affirms its own existence, must exist beyond physical substance and possesses the capacity to gain broader knowledge (Ibn Sīnā 2002, 233–35). Consciousness must exist first in order that it may become self-consciousness (self-awareness), which is aware of being in the surrounding environment.

Ibn Sīnā’s philosophy places the soul (nafs) at the core of human existence, an immaterial (spiritual) and eternal substance that uniquely inhabits the body. He divides the soul into three interdependent and hierarchical levels: the vegetable soul (nafs nabātiyyah), responsible for basic functions such as nutrition, growth, and reproduction; the animal soul (nafs ḥayawāniyyah), which adds sensory perception and movement; and the rational soul (nafs nāṭiqah), which is at the top of the hierarchy of souls and unique to humans, encompassing the faculties of intellect (‘aql) and will (Ibn Sīnā 1997, 54–70). For Ibn Sīnā, artifacts categorically lack consciousness because they lack souls (Ibn Sīnā 2017, 13–14; Ṣadrā 2002, 153–56). AI, as an artifact built on nonliving hardware, denies the existence of a soul and therefore lacks this self-awareness. This highlights a core difference in consciousness and challenges the purely materialist view of AI (Alawijeh and Al-Mesilini 2025). From the perspective of Ibn Sīnā’s thought, AI is an artificial entity whose principle of motion is external and uniform, so it does not have a “container” for self-awareness, which is a prerequisite for understanding.

 2. Understanding

According to Ibn Sīnā (2002, 237–43), human understanding involves abstraction (tajrīd) from particular forms to universal concepts. Through the faculty of estimation (wahm), humans grasp “non-sensory meanings” (ma‘nā)—such as goodness or bothness—that cannot be perceived by the external senses (Rahman 1981, 30–31).

Modern AI, on the other hand, operates by manipulating symbols and statistical patterns without understanding the essence or intentionality behind them. Natural intelligence’s understanding is substantial (changing the cognizing soul), while AI’s “understanding” is merely an accidental adjustment of data positions on hardware.

Through the “Chinese Room” thought experiment, John Searle concluded that machines cannot have “understanding,” regardless of how intelligent or humanlike the program is. They do not truly “understand” or “believe” anything in the way humans do because they lack real intentionality (Searle 1980).

Another perspective can be explored through Ibn Sīnā’s concept of knowledge. He regards knowledge as an essential attribute of its subject matter (mawḍū‘). According to Ibn Sīnā, the subject matter of a specific branch of knowledge or science is the central problem that defines that field. The subject matter serves as the genus with the broadest scope, encompassing all the particular issues addressed within that science. What distinguishes one science from another is the difference in their subject matters. If the subject matter of a particular science is discussed in another field, it becomes a problem within that other science. The properties or predicates (essential accidents) attached to the subject matter become the specific problems examined within a given science. For example, the subject matter of natural science (ṭabī‘iyyāt) is bodies (ajsām, sing. jism) considered from the perspective of their motion and rest. The issues addressed in natural science are the accidents that befall bodies in terms of movement or stillness. Thus, concepts such as jism (body) as being, substance, matter, and form are not the focus of natural science, as these belong to a different domain. When the subject matter of a science is not addressed within that science itself, it must be examined by a higher science—namely, metaphysics. Meanwhile, the problems discussed in natural science are predicates (accidents) related to moving or stationary bodies (ajsām, sing. jism), such as mineral science, composites, plant science, animal nature, the study of the soul, senses, sense objects, chemistry, and other issues that remain within the scope of natural science (Ibn Sīnā 1983, 4–10).

Mathematics, which encompasses algorithms and computations, is concerned with the study of quantity (kam). However, the task of proving the very existence of “quantity” lies beyond the scope of mathematics itself; this is a question reserved for metaphysics. If mathematics were to attempt to prove the existence of quantity by relying on the concept of quantity, it would result in circular reasoning, which is logically untenable (Ibn Sīnā 1983, 4–10). AI, operating within the boundaries of computational cognition and algorithms, is limited to processing the properties (essential accidents) of quantities as data. In contrast, humans possess the capacity for metaphysical or philosophical thought, enabling them to understand the existence and nature of data—an ability that extends beyond the current capabilities of AI.

This aligns with Penrose’s view that computational algorithms, no matter how powerful, cannot produce genuine understanding. Penrose uses Gödel’s incompleteness theorems to support this. These theorems state that in any sufficiently complex mathematical system, there are true statements that cannot be proven within the system. Penrose argues that the human mind can grasp and prove these statements, suggesting that human understanding involves non-algorithmic processes that computers cannot replicate (Penrose 2002, 106–9; Penrose and Severino 2022, 33–34).

Human understanding allows us to “see” truths that lie beyond the reach of any formal system. Therefore, if human intelligence involves the ability to understand these non-computable truths, then intelligence itself cannot be algorithmic (Penrose and Severino 2022, 41–43). AI merely borrows human intelligence through algorithms. AI is a functional extension of human reason, not an independent intellectual agent. Without the capacity for understanding rooted in consciousness, AI will always be a pseudo-“intelligence”—a simulation without a subject.

Supported by Penrose’s thinking, AI ontology, from Ibn Sīnā’s perspective, positions AI as an entity that will never attain true intelligence. Because AI is ontologically purely algorithmic matter, it is cut off from the chain of consciousness–understanding interdependence. AI remains a ṣinā‘ī (artificial) entity—a highly sophisticated functional illusion that mimics the results of human intelligence without ever possessing the essence of understanding itself. This also refutes the possibility of strong AI occurring.

 3. Intelligence

Consequently, what appears as cognition in AI is better understood as formal manipulation rather than intellectual apprehension. In Ibn Sīnā’s metaphysical framework, intelligence represents the actualization of the rational soul’s potential. The rational soul’s highest goal is sa‘ādah (true happiness), which is achieved through self-perfection and the actualization of one’s intellectual potential. This view provides a clear teleology to ethics, where noble character (akhlāq) is not merely a set of rules but an essential means to achieving spiritual perfection (Ibn Sīnā 2024, 5). This teleological framework distinguishes natural from artificial entities. Natural beings possess intrinsic ends; artifacts have only extrinsic purposes assigned by makers.

This perfection of the rational soul occurs through a gradual process, which Ibn Sīnā describes through four stages of human reason. This process begins with ‘aql hayūlānī (material intellect), which is the pure potential for thought, similar to a blank slate ready to be written on. The next stage is ‘aql bi al-malakah (intellect in talent), in which the intellect has acquired the “first intelligible” or indisputable fundamental principles, like the axioms in mathematics. The intellect then develops into ‘aql bi al-fi‘l (actual intellect), which possesses universal knowledge but does not actively think about it. The culmination is ‘aql mustafād (acquired intellect), a state of full actualization in which the intellect actively and consciously considers the intelligible it has accessed (Ibn Sīnā 2022, 34–37).

The human rational soul (nafs nāṭiqah) that begins from ‘aql hayūlānī, which is a potential intellect, has two paths to reach a high level (‘aql mustafād), namely fikr (reason) and ḥads (intuition), owned by people whose intellect is already very strong. If fikr still uses rational–logical procedures, ḥads is able to produce the middle term in a syllogism without compiling a major premise and a minor premise (Ibn Sīnā 1997, 300–4).

This process of intellectualization does not occur spontaneously but is enabled by the external intervention of ‘aql fa‘āl (active intellect), the tenth intellect in Ibn Sīnā’s cosmic hierarchy of emanations. The active intellect functions as the “giver of form” (wāhib al-ṣuwar), that is, the eternal and perfect metaphysical entity responsible for actualizing the human intellect (Ibn Sīnā 1997, 332–33). Active intellect illuminates human intellect, enabling it to move from potentiality to actualization and attain universal intelligibility. This process is called “contact” (ittiṣāl) or “divine effusion” (fayḍ ilāhī). The active intellect eternally considers all intelligible, which then flow into the human intellect that is ready and actively seeking them (Rahman 1981, 68–69).

To understand how the human mind attains universal knowledge, Ibn Sīnā outlined the role of internal faculties tasked with processing sensory data before the rational mind acts (Sīnā 2002, 239–41). Ibn Sīnā believes perception involves both external and internal senses. Perceptions from external and internal senses produce particular knowledge. The external senses are touch, smell, taste, sight, and hearing. The internal senses, located in the brain, include:

a. common sense (ḥiss mushtarak): integrates sensory data

b. imagery/fantasy (khayāl): retains sensory images

c. memory (hāfiẓah/dhākira): preserves meanings and judgments

d. estimation (wahmiyah): evaluates meanings and forms opinions

e. imaginative faculty (mutakhayyila or mufakkirah): analyzes, synthesizes, and manipulates forms to derive meaning (Sīnā 2002, 239–41; Silva and Yrjönsuuri 2013, 101).

Estimation captures the non-sensory meaning of partial sensory objects, such as the faculty that determines that wolves should be avoided (the reception of ma‘ānī). It is located in the back of the brain’s mid-cavity. Estimation perceives meaning in particular things such as love and respect for a person, fear of an animal, pain, friendship, enmity, and so on (Rahman 1981, 30–33).

Mutakhayyila is a kind of “processor” that processes forms obtained from outside that are captured by the senses, while mufakkirah is a kind of “processor” that processes meanings that are universal in nature, not obtained from outside but used by the human intellect. The mufakkirah is unique to humans and allows for rational thought, judgment, and insight. Through this faculty, humans can process abstract ideas beyond sensory perception (Sīnā 2002, 240–41). The combination of these faculties allows the human process of intellectualization to culminate in tajrīd (abstraction), where the human intellect extracts the essence or universal form of matter for thought.

Ibn Sīnā highlights the importance of imagination in the creative process. He sees imagination as the ability that allows humans to form images and ideas not directly perceived. This imaginative power is essential for both artistic and scientific creativity. Ibn Sīnā describes creativity as a process involving both intellect and imagination. Intellect provides the structure and principles, while imagination brings these principles to life by forming new and unique combinations. This interaction between intellect and imagination allows humans to create art, literature, and scientific theories. Intellect plays a vital role in creative endeavors, enabling humans to grasp complex concepts and generate new ideas or innovations (Sīnā 2002, 246–47).

From the aforementioned description, Ibn Sīnā’s theory of intelligence can be explained as the transformation of knowledge acquisition from its lowest degree to the highest through a universal agency (the Active Intellect) and through ethical and spiritual practices that shape the function of human intelligence. The term intelligence has different scopes between humans and AI. In humans, it encompasses reason, intuition, understanding, wisdom, moral conscience, analytic, and aesthetic judgment, whereas in AI, it only refers to the analytic function of computation.

Although AI may appear to be able to perform similar processes through feature extraction and pattern recognition, these processes are very different. Abstraction in AI is a process in which a model learns to identify and discard information deemed “irrelevant” based on a predetermined goal (Lin 2024). This abstraction is built from specific data and statistics with validity in the training domain but that can fail in other domains. Because the model only generalizes to the extent that patterns emerge in the data, it does not produce universal knowledge, only local representations useful for specific tasks. The model can change if the data and parameters change.

AI is a fundamentally closed cognitive system, operating on data and algorithms designed by humans. Its source of knowledge is limited to curated datasets, which are essentially collections of partial and ephemeral empirical information (Lin 2024). AI has no Active Intellect or eternal and perfect external entity to relate to. Consequently, AI can never achieve true ‘aql mustafād, or pure universal understanding, in Ibn Sīnā’s sense.

AI operates primarily at the levels of exploratory and combinational creativity. AI can explore existing datasets and combine elements, but it lacks the capacity for transformational creativity, which requires intuition, inspiration, and the ability to transcend existing paradigms (Sautoy 2019, 7–15). Data cannot produce output until it is converted into code and algorithms. It suggests that true creativity requires something beyond complex computation (Sautoy 2019, 10). AI is a tool for creativity, not creativity itself. While AI can assist in generating ideas and exploring possibilities, it cannot replace the human capacity for genuine, transformative creativity that produces groundbreaking innovations, often arising in rare “eureka” moments that change the world (Jordan 2019). For example, AI can compose music by studying a large number of existing pieces and spotting patterns, but it does not create something from true inspiration or emotional depth (Newblom 2026).

It seems that AI can only imitate and amplify sensory functions and some inner senses related to sensory or material objects. Meanwhile, objects from within the mind, such as dreams, feelings (such as fear or love), and inspiration, cannot be captured by AI. While AI imitates the structure of the material brain and some cognitive functions, it has not yet replicated the full complexity of the human brain, let alone the human intellect. No matter how sophisticated brain-boosting technology or interfaces support nervous system function, it will never reach human intellect.

What about deep learning? Machines seem to “learn” on their own. The “learning” process in AI is just a change in numerical values (weights) in memory. For Ibn Sīnā, this is not the “achievement of knowledge” by the machine but rather accidental movement (ḥarakah ‘araḍiyyah). These weight values are accidental quantities changed by the optimization algorithm. The machine does not become “wiser,” it simply becomes more precisely configured to map inputs to specific outputs. Learning in humans is the transformation of the soul from potential to actual, while AI “learning” is simply the adjustment of positional and quantity accidents in silicon by external agents (initial data and algorithms).

 Decision-Making: Artificial Intelligence Recommendations and the Nuance of Human Wisdom

AI offers decision-making support by leveraging extensive data and algorithms. Generally, larger datasets lead to more accurate AI-driven recommendations. AI excels quickly and accurately, processing vast amounts of data, identifying patterns, and making decisions based on statistical analysis. It can analyze data and reach decisions much faster than humans, and free from fatigue, a real advantage in time-sensitive situations (Martin 2023).

AI can function as a civilizational instrument that relieves human cognition from excessive preoccupation with particular repetitive and instrumental tasks. Much of modern human life is absorbed by managing fragmented data, procedural decisions, and technical operations that burden the brain while leaving little space for the higher operations of intellect (ʿaql). In this context, AI can serve as a form of cognitive offloading, assisting the brain’s operational functions and allowing human intelligence to reorient itself toward theoretical reflection, ethical judgment, and the pursuit of truth.

For Ibn Sīnā, the highest function of human intelligence lies not in mechanical problem-solving but in the actualization of the rational soul through theoretical understanding and practical wisdom. When human beings are liberated from excessive engagement with the accidental and the particular, they are better positioned to cultivate ʿaql naẓarī (theoretical intellect), which apprehends universals and seeks truth, and ʿaql ʿamalī (practical intellect), which guides ethical action (Ibn Sīnā 2022, 31–33; Marmura 2005, 2).

In this sense, AI—precisely because it lacks consciousness, understanding, and moral intention—can become an assistant tool rather than a rival, provided it remains subordinated to human intellectual and moral governance. The ethical task, therefore, is not to endow AI with moral agency but to ensure human beings retain and strengthen their own capacity for wisdom (ḥikmah), thereby guiding technological power toward the realization of the good (khayr) and the true (ḥaqq).

 Conclusion

This article has examined AI through philosophical dimensions drawing on the comprehensive intellectual framework of Ibn Sīnā. Within Ibn Sīnā’s ontology, his distinction between natural (ṭabī‘ī) and artificial (ṣinā‘ī) entities provides a powerful lens through which to understand AI’s fundamental nature. AI, however sophisticated, is properly classified as a ṣināʿī (artificial) entity, an accidental configuration of matter whose operation depends entirely on external agents, material substrates, and human intentionality. As such, AI lacks the ontological conditions required for genuine intelligence in the classical sense.

The analysis demonstrated that human intelligence, in Ibn Sīnā’s philosophy, is inseparable from the rational soul (nafs nāṭiqah), self-awareness, abstraction, and the teleological movement from potentiality to actuality. Intelligence is not reducible to computation, pattern recognition, or statistical optimization but is the actualization of an immaterial subject capable of understanding universals and orienting itself toward truth and perfection. AI systems, by contrast, operate exclusively within the domain of quantity, algorithmic manipulation, and accidental change. Their apparent “learning” and “creativity” remain functional simulations rather than manifestations of understanding or consciousness.

From this perspective, claims concerning strong AI, machine consciousness, or artificial moral agency are metaphysically untenable. No increase in computational complexity or data scale can bridge the ontological gap between artificial systems and natural intellectual beings. AI does not possess an internal principle of life, will, or intentionality and therefore cannot be a bearer of moral responsibility. Ethical accountability must remain fully with human agents—designers, developers, institutions, and users—who employ AI as an instrumental extension of human reason.

Ultimately, Ibn Sīnā’s philosophy offers a critical and constructive framework for this task, reaffirming the uniqueness of human intelligence while resisting both technological reductionism and anthropomorphic projections onto machines. In the ongoing dialogue between religion and science, this Avicennian perspective underscores a central insight: technological power advances with ease, but wisdom advances only insofar as human beings remain oriented toward truth (ḥaqq) and the good (khayr).

 Competing Interests

The authors declare that they have no competing interests.

 Adamson, Peter. 2016. Philosophy in the Islamic World. Oxford: Oxford University Press. Alawijeh, Mustafa ‘Azizi, and Seyed Zuhair Al-Mesilini. 2025. “A Comparative Study of Artificial and Natural Intelligence from Ibn Sina’s Perspective.” Journal of Ecohumanism 4 (1): 1. 10.62754/joe.v4i1.6082. Arif, Syamsuddin. 2007. “Ibn Sina’s Idea of Nature and Change.” Afkar: Jurnal Akidah Dan Pemikiran Islam 8 (1): 111–40. Belo, Catarina. 2009. “Essence and Existence in Avicenna and Averroes.” Al-Qanṭara 30 (2): 403–26. 10.3989/alqantara.2009.v30.i2.84. Coeckelbergh, Mark. 2020. AI Ethics. Cambridge, MA: MIT Press. Copeland, B. J. 2026. “Artificial Intelligence (AI).” In Encyclopædia Britannica. Accessed February 13. https://www.britannica.com/technology/artificial-intelligence. Cycleback, David. 2018. Philosophy of Artificial Intelligence. London: Bookboon. Dreyfus, Hubert L. 1992. What Computers Still Can’t Do: A Critique of Artificial Reason. Cambridge, MA: MIT Press. Google Cloud. n.d. “What Is Artificial Intelligence (AI)?” Accessed February 14, 2026. https://cloud.google.com/learn/what-is-artificial-intelligence. Haenlein, Michael, and Andreas Kaplan. 2019. “A Brief History of Artificial Intelligence: On the Past, Present, and Future of Artificial Intelligence.” California Management Review 61 (4): 5–14. 10.1177/0008125619864925. Ibn Sīnā, Abū ʿAlī al-Ḥusayn ibn ʿAbd Allāh. 1960. Al-Ilāhiyyāt Fī Kitāb al-Shifā. Edited by Qanwātī, Sa‘īd Zāyd, and Ibrāhīm Madkūr. Cairo: Al-Hayʾah al-Miṣriyyah al-ʿĀmmah lil-Kitāb. Ibn Sīnā, Abū ʿAlī al-Ḥusayn ibn ʿAbd Allāh. 1983. Al-Thabī’iyyāt Min Kitāb Al-Shifā. Cairo, Egypt: Markaz Tahqīq Al-Turāts Ibn Sīnā. Ibn Sīnā, Abū ʿAlī al-Ḥusayn ibn ʿAbd Allāh. 1997. Al-Nafs Min Kitāb al-Shifā’. Edited by Ḥasan Zādeh Āmulī. Qom, Iran: Markaz al-Nasyr-Maktab al-I‘lam al-Islāmī. Ibn Sīnā, Abū ʿAlī al-Ḥusayn ibn ʿAbd Allāh. 2002. Al-Isyārāt Wa al-Tanbīhāt. Edited by Mujtabā al-Zāri‘ī. Beirut: Mu’asseseh Bustān Kitāb. Ibn Sīnā, Abū ʿAlī al-Ḥusayn ibn ʿAbd Allāh. 2017. Risālah Fī Māhiyat Al-‘Ishq. London: Hindawi. Ibn Sīnā, Abū ʿAlī al-Ḥusayn ibn ʿAbd Allāh. 2022. Aḥwāl Al-Nafs: Risālah Fī an-Nafs Wa Baqā’ihā Wa Ma‘ādihā and Thalāth Ar-Rasāil Fī an-Nafs. Translated by Achmad Syahid. Jakarta, Indonesia: Turos. Ibn Sīnā, Abū ʿAlī al-Ḥusayn ibn ʿAbd Allāh. 2024. “Fī ‘Ilm al-Akhlāq.” In Tisʿ Rasāʾil Fī Al-Ḥikmah Wa-al-Ṭabīʿiyyāt. Dār al-ʿArab. Translated by Deni Ramdani. Fins. Jordan, Jamila. 2019. “‘The Creativity Code’: Is AI Taking Over Creative Industries?” Arts Management and Technology Laboratory. November 7, 2019. https://amt-lab.org/reviews/2019/11/the-creativity-code-is-ai-taking-over-creative-industries. Lin, Yu-Jen. 2024. “Understanding Abstractions in Neural Networks.” Towards Data Science. May 14, 2024. https://towardsdatascience.com/understanding-abstractions-in-neural-networks-22cc2cd54597/. Marmura, Michael E. 2005. Avicenna: The Methaphisic of The Healing. Provo, UT: Brigham Young University Press. Martin, Jeff. 2023. “AI vs Human Intelligence: Accuracy and Error Rates.” Future Savant. December 1, 2023. https://futuresavant.com/ai-vs-human-intelligence-accuracy-and-error-rates/. McGinnis, Jon. 2010. Avicenna. Oxford: Oxford University Press. McGinnis, Jon. 2025. “Ibn Sina’s Natural Philosophy.” In The Stanford Encyclopedia of Philosophy, Spring 2025, edited by Edward N. Zalta and Uri Nodelman. Metaphysics Research Lab, Stanford University, March 20, 2025. https://plato.stanford.edu/archives/spr2025/entries/ibn-sina-natural/. Mohaddes, Alireza. 2020. “An Introduction to the Value of Technology from the Perspective of the Islamic Philosophy.” Ma‘rifat Farhangi Ejtemaei 11 (4): 29–46. Newblom, Michelle. 2026. “AI vs Human: Creativity, Abilities, and Skills (Which Is Better?).” Fiverr. August 2, 2026. https://www.fiverr.com/resources/guides/business/ai-vs-human. Norman, Nurul Ain. 2021. “Ibn Sina’s Theory of the Soul: A Taxonomy of Islamic Education.” ICR Journal 12 (2): 275–89. 10.52282/icr.v12i2.867. Penrose, Roger. 2002. The Emperor’s New Mind: Concerning Computers, Minds, and the Laws of Physics. Oxford: Oxford University Press. Penrose, Roger, and Emanuele Severino. 2022. “A Dialogue on Artificial Intelligence versus Natural Intelligence.” In Artificial Intelligence Versus Natural Intelligence, edited by Roger Penrose, Emanuele Severino, Fabio Scardigli, et al. Cham, Switzerland: Springer International Publishing. 10.1007/978-3-030-85480-5_2. Popova, Biliana. 2020. “Islamic Philosophy and Artificial Intelligence: Epistemological Arguments.” Zygon: Journal of Religion and Science 55 (4): 4. 10.1111/zygo.12651. Rahman, Fazlur. 1981. Avicenna Psychology: An English Translation of Kitāb al-Najāt, Book II, Chapter VI. Westport, CT: Hyperion Press, Inc. Ṣadrā, Mullā. 2002. Al-Ḥikmah al-Muta‘āliyah Fī al-Asfār al-‘Aqliyyah al-Arba‘ah. Vol. 7. Beirut, Lebanon: Dār Iḥyā’ al-Turāth al-‘Arabī. Sautoy, Marcus du. 2019. The Creativity Code: Art and Innovation in the Age of AI. Cambridge, MA: Harvard University Press. Searle, John. 1980. “Minds, Brains, and Programs.” Behavioral and Brain Sciences 3 (3): 417–57. Shareef, Omar. 2023. “The Realizability of Strong AI: A Deeper Look from an Islamic Perspective.” Bachelor’s thesis, Harvard College. Silva, José Filipe, and Mikko Yrjönsuuri. 2013. Active Perception from History of Philosophy. Cham, Switzerland: Springer. Tulenko, Abigail. 2024. “What Philosopher Ibn Sina Can Teach Us about AI.” Scientific American, April 18, 2024. https://www.scientificamerican.com/article/what-philosopher-ibn-sina-can-teach-us-about-ai/. Vallor, Shannon. 2016. Technology and the Virtues: A Philosophical Guide to a Future Worth Wanting. Oxford: Oxford University Press. Wheeler, Michael. 2005. Reconstructing the Cognitive World. Cambridge, MA: MIT Press. Zarepour, Mohammad Saleh. 2019. “Avicenna’s Philosophy of Mathematics.” PhD diss., University of Cambridge.