What the Encyclical Cannot Do Alone: AI, Interiority, and the Work Ahead for Catholic Psychology

In Paragraph 99 of Magnifica Humanitas, Pope Leo XIV writes that artificial intelligence systems 'do not undergo experiences, do not possess a body, do not feel joy or pain, do not mature through relationships and do not know from within what love, work, friendship or responsibility mean.' They may simulate empathy, he continues, but 'they do not understand what they produce, for they lack the affective, relational and spiritual perspective through which human beings grow in wisdom.'[^1]

The encyclical offers no philosophical argument for this position, and that absence has been noted. Christopher Olah, co-founder of Anthropic, spoke at the Vatican event on May 25, 2026, describing internal AI structures that 'mirror results from human neuroscience' and 'functional' states resembling joy, fear, and grief. He pressed for continued discernment on whether those findings carry deeper significance.[^1] The tension is real: the Pope asserts a boundary; a leading AI researcher finds that boundary harder to locate than expected.

What the encyclical does, and what makes it worth taking seriously, is force a definition. If one rejects the Pope's claim, one must say what consciousness is such that an AI might possess it. That is a harder task than it appears.

Catholic anthropology locates consciousness not in functional complexity but in the soul's capacity for self-transcendence — a turning toward truth, goodness, and God that is not reducible to information processing. Aquinas distinguishes the intellective soul from every material operation precisely because the intellect grasps universals that no organ, however refined, can produce.[^2] A system that generates probabilistic text outputs, however sophisticated, operates entirely within the order of instrumental causation. It has no interiority in the Thomistic sense: no esse that is its own, no act of understanding that reaches beyond the sign to the thing signified.

This is not a claim about current AI capability or future AI development. It is a claim about category. Consciousness, on the Catholic account, is not a threshold that engineering can cross by accumulating scale. It is the condition for moral agency, genuine suffering, genuine joy, and the kind of relational encounter that constitutes human life at its depth. An entity without interiority cannot be wronged in the way a person can be wronged, cannot love in the way a person can love, and cannot be a subject of care in the way that pastoral and clinical practice presuppose. Vitz, Nordling, and Titus (2020) ground exactly this claim in the CCMMP's account of the person: the human being is a unity of body and soul whose dignity flows from a created nature ordered toward God, not from any measurable functional property.[^3]

Viktor Frankl's logotherapy, developed from his experience in the concentration camps, argued that the search for meaning is the primary human motivation — not pleasure, not power, but meaning grasped by a subject who can suffer its absence. That grasp requires precisely what AI lacks on Leo's account: an interior life capable of encountering reality as reality, not merely processing inputs and generating outputs.

The commentary in the National Catholic Register frames Leo's statement as an invitation to further discernment rather than a foreclosure of inquiry.[^1] That framing is apt. Discernment, in the Catholic tradition, is a practiced capacity for reading reality accurately — not skepticism, and not credulity, but attentiveness trained by both reason and grace. Applied to artificial intelligence, discernment asks what is actually present when a system produces a response that resembles understanding. The encyclical's answer is that resemblance is not identity, and that confusing the two carries costs that fall on human persons.

The question of AI consciousness is, at bottom, a question about what human beings believe themselves to be. If consciousness is a sufficiently complex pattern of information processing, the human person is a matter of degree rather than kind. If consciousness is the act of a soul that exceeds its material conditions, the human person occupies a category no engineering project can replicate. The Catholic tradition holds the second position, and holds it not as a defensive reflex against technology but as the foundation of everything it has said about dignity, moral agency, and the irreducible worth of the person.

Catholic psychologists and neuroscientists: where the work now lies

Magnifica Humanitas names the boundary between human consciousness and artificial processing. What it does not do — and what Leo's encyclical cannot do alone — is supply the detailed empirical and philosophical account needed to sustain that boundary under pressure from researchers like Olah. That is work for Catholic psychologists and neuroscientists, and it is work the tradition has begun but not finished.

The most developed starting point is in Thomistic cognitive psychology. Benjamin Suazo's account of the cogitative sense — the vis cogitativa that Aquinas identifies as the faculty through which the human being grasps the particular as meaningful, not merely as data — offers a precise locus for what AI processing lacks.[^4] The cogitative sense integrates sensory perception with intellectual judgment in a way that is irreducibly personal: it is the faculty through which a person recognizes this individual as a friend, this situation as one calling for courage, this loss as genuinely mine. No large language model exercises anything analogous because the cogitative sense is the act of a unified subject, not a pattern-matching operation over token sequences. Catholic neuropsychology needs to develop this distinction in dialogue with current cognitive science — showing not only what the scholastic account claims but where the neuroscientific evidence either coheres with or strains against it.

A second area where development is needed is the neuroscience of consciousness itself. The neurobiologist Alberto Carrara, writing from the Pontifical Athenaeum Regina Apostolorum, has noted that the term 'consciousness' is polysemous across neuroscience, psychiatry, and philosophy in ways that generate persistent confusion.[^5] What AI researchers identify as 'functional states' mirroring joy or fear are states defined operationally — by their causal role in system behavior. What Catholic anthropology means by consciousness is the subjective, first-person character of experience: what Thomas Nagel, in a text Carrara cites, called the irreducible fact that there is 'something it is like' to be a conscious creature. These are not the same question, and conflating them allows a plausible empirical finding — that AI systems have internal states with functional properties — to appear to bear on a philosophical question it does not actually address. Catholic neuroscientists are positioned to make this distinction with precision, and to do so in venues where AI researchers will encounter it.

The Antiqua et Nova note on artificial intelligence from the Dicastery for the Doctrine of the Faith made this distinction explicit, arguing that AI's ability to produce responses 'associated with the human intellect' does not constitute the ability to think, because the evaluation of AI performance is methodologically reductive: it measures outputs, not the act of understanding that in human beings generates those outputs.[^6] That argument needs elaboration from within cognitive neuroscience. The act of understanding, as the Christian philosophical tradition has described it, involves a reaching beyond the sign to the thing signified — an intentional directedness toward reality that is not reducible to any physical or computational process, however complex. Whether current neuroscience can identify neural correlates of this intentionality without reducing it is precisely the question Catholic neuroscientists should be pressing.

Vitz, Nordling, and Titus situate this within the CCMMP's account of the person as a created being whose cognitive life includes both the material operations of sensation and the immaterial operations of intellect — a unity that secular cognitive science typically dissolves by treating all cognition as computation.[^3] The clinical implication is not abstract: how practitioners understand consciousness shapes how they understand suffering, moral responsibility, and the capacity for growth through relationship. A therapeutic model that treats the patient as a complex information-processing system will systematically miss what the cogitative sense and the intellective soul are doing in the patient's experience of their own life. Catholic clinical psychologists — especially those trained in the integration represented by the CCMMP — are among the few practitioners equipped to hold both the neuroscientific and the anthropological register simultaneously.

What the encyclical opens, then, is not a settled conclusion but a research program. The philosophical claim that AI lacks consciousness because it lacks interiority requires Catholic thinkers to say with greater precision what interiority is, where in the person it is exercised, what its neural correlates are or are not, and why the functional states Olah identifies do not constitute it. That program runs through Suazo on the cogitative sense, through Carrara on the polysemy of consciousness, through the CCMMP's account of the soul-body unity, and through the ongoing conversation between Thomistic anthropology and cognitive neuroscience. It will not be completed by a single encyclical. But Magnifica Humanitas has made it impossible to defer.

References

[^1]: Joshua Hochschild, 'No, AI Isn't Conscious. But Saying So Invites Further Discernment,' National Catholic Register, June 10, 2026. https://www.ncregister.com/commentaries/no-ai-isn-t-conscious-discernment

[^2]: Thomas Aquinas, Summa Theologiae (1265–1273), I, q. 79, a. 2; I-II, q. 22, a. 1–4.

[^3]: Paul C. Vitz, Craig S. Titus, and William J. Nordling, A Catholic Christian Meta-Model of the Person: Integration with Psychology and Mental Health Practice (Divine Mercy University Press, 2020), ch. 4.

[^4]: Benjamin Suazo, on the cogitative sense and its role in moral perception; see also J. A. Tellkamp, 'Vis aestimativa and vis cogitativa in Thomas Aquinas's Commentary on the Sentences,' The Thomist, 76(4), 611–640.

[^5]: Alberto Carrara, L.C., 'Coscienza o coscienze? Aspetti antropologici e risvolti etici della ricerca neuroscientifica sugli stati di coscienza,' Gruppo di Neurobioetica, Ateneo Pontificio Regina Apostolorum, Roma.

[^6]: Dicastery for the Doctrine of the Faith, Antiqua et Nova: Note on the Relationship Between Artificial Intelligence and Human Intelligence, 2025, §§ 12–14.