When Algorithms Govern the Soul: China's AI Religious Surveillance and the Catholic Vision of Human Dignity

China's National Religious Affairs Administration is deploying generative AI to filter religious content across social media, targeting any expression that deviates from state-approved doctrine. This development arrives as Pope Leo XIV's encyclical Magnifica Humanitas reframes the global conversation about technology, conscience, and the irreducible dignity of the human person. The contrast is both geopolitically and anthropologically clarifying.

June 23, 20266 min read
When Algorithms Govern the Soul: China's AI Religious Surveillance and the Catholic Vision of Human Dignity

A new academic centre at Southeast University in Jiangsu province, China, formerly the National Nanking University, has published findings that bear directly on the intersection of faith, psychology, and human flourishing. The Centre for Intelligent Computing of Religious Information in Asia, Africa, and Europe, led by Professor Wang Qilong and researcher Sun Xiaoli, is not a theoretical enterprise. As Asia News and ZENIT reported on June 20, 2026, it is an applied institution whose explicit purpose is to provide Chinese public administration with AI tools capable of filtering religious content at a scale and precision that previous censorship methods could not achieve.

Prior approaches to online religious governance in China relied on blocking specific sensitive words. The generative AI models under development at Southeast University operate differently: they analyze context, detect nuance, and identify any content that does not conform to the political vision of religion under what the state calls new-era socialism. The WeChat account Weiyan Zongjiao, the official voice of China's National Religious Affairs Administration, published the academic article as a statement of institutional direction. Religious expression in China is being subjected to computational governance, and the architecture being built is designed to be comprehensive.

The encyclical and the algorithm: two visions of the human person

The timing of this disclosure is not incidental. Pope Leo XIV's encyclical Magnifica Humanitas has positioned the Catholic tradition as a serious interlocutor in debates about technology, conscience, and what it means to be a person at all. That a state apparatus is simultaneously deploying AI to suppress precisely the kind of interior religious life that Magnifica Humanitas seeks to protect creates a tension that is both geopolitical and deeply anthropological.

At the center of Catholic anthropology is a claim that is not merely doctrinal but has extensive psychological grounding: the human being is a unified whole, a person whose spiritual, psychological, and relational dimensions are inseparable. The Catholic Christian Meta-Model of the Person, developed by Vitz, Nordling, and Titus, does not treat faith as a private preference that can be safely bracketed from wellbeing. It treats the relationship between the person and transcendence as constitutive of psychological health.[^1] This is not sentiment. Research in positive psychology, including work drawing on self-determination theory, has consistently found that meaning-making frameworks rooted in religious and spiritual life contribute to psychological coherence and the capacity to endure suffering without fragmentation.

What the Southeast University centre is engineering is, in this light, not merely a censorship system. It is a mechanism for severing the connection between persons and their own interior lives, mediated through digital communication.

Religious suppression as a mental health question

Clinical and research traditions in mental health have long recognized that interference with religious identity produces measurable psychological harm. Studies on religious minorities living under authoritarian governance document elevated rates of anxiety, moral injury, and what researchers have called existential dislocation — a state in which the person cannot access the meaning-structures that normally sustain psychological equilibrium. When digital environments are engineered to make authentic religious expression invisible, unsearchable, and unreachable, the effect is not simply political. It is therapeutic deprivation at scale.

The Catholic approach to mental health takes this seriously because it begins from a different premise than purely secular models. The person is not a brain producing experiences. The person is a being oriented toward truth, beauty, goodness, and ultimately toward God, and this orientation is not incidental to health but central to it.[^1] When a state apparatus uses generative AI to ensure that only algorithmically approved interpretations of religious experience can circulate, it is not governing religion. It is attempting to govern the soul — to determine what the interior life of millions of persons is permitted to be.

Self-determination theory, developed by Edward Deci and Richard Ryan, identifies autonomy, competence, and relatedness as fundamental psychological needs whose frustration produces not simply dissatisfaction but genuine ill-being.[^2] Religious practice, when freely chosen and authentically expressed, satisfies all three. It provides agency in the encounter with ultimate meaning, competence in the navigation of suffering and mortality, and relatedness within communities of shared belief. An AI system designed to eliminate non-compliant religious content is, from this framework, an instrument of systematic need-frustration.

The architecture of spiritual erasure

Southeast University houses a well-established School of Foreign Languages, described on its own website as a bridge between cultures, with the first English teacher training programme in China dating to 1917. The same institution that positioned itself for over a century as a mediator between Chinese and global intellectual life now hosts a centre whose purpose is, in significant part, to insulate Chinese religious life from foreign influence. That phrase, as used in the National Religious Affairs Administration's framing, covers any religious interpretation originating outside state-sanctioned doctrine.

This is the language of closure. The Catholic tradition is constitutively open. Its intellectual history is one of sustained engagement with whatever is true wherever truth is found — from Aquinas's integration of Aristotelian philosophy to the dialogues of the Second Vatican Council with modernity. The Meta-Model of the Person that informs serious Catholic mental health work is not a closed system of approved thoughts. It is a framework for encountering reality, including the reality of suffering, complexity, and genuine encounter with the other, with intellectual honesty and spiritual depth.[^1]

What the data cannot capture

The very scale of this enterprise — a dedicated academic centre at a major university, AI models built specifically for social media religious content, an official government publication promoting the project — testifies to the significance of what is being targeted. Religious faith, freely practiced and authentically expressed, is understood by those seeking to control it as a genuine source of human agency that cannot be reduced to political compliance.

The question that Magnifica Humanitas raises, and that the Southeast University centre inadvertently answers, is not whether AI can be used to govern religious life. Plainly it can, at least in terms of shaping digital environments. The deeper question is what vision of the human person any given use of technology presupposes and serves.

The Catholic tradition answers this question with a claim that is both ancient and urgently contemporary: the person is irreducible. No algorithm, however sophisticated, can produce what freely chosen faith and authentic spiritual community produce in the interior life of the human being. The therapeutic work of supporting human flourishing — whether in clinical settings, pastoral contexts, or broader cultural engagement — begins from exactly this premise.

As AI becomes an increasingly pervasive feature of the environments in which persons seek meaning, connection, and healing, the anthropological foundations of that work matter more, not less. The contrast between an encyclical that affirms the magnificent humanity of the person and a censorship system designed to engineer compliant religious subjectivity is a clarifying moment about what is genuinely at stake in the intersection of technology, faith, and the interior life.

Source: ZENIT News / Asia News, June 20, 2026. Original reporting by Asia News, Milan.

References

[^1]: Craig Steven Titus, Philosophy of Mental Health (Divine Mercy University Press, 2020); in Paul C. Vitz, William J. Nordling, and Craig Steven Titus, A Catholic Christian Meta-Model of the Person: Integration with Psychology and Mental Health Practice, pp. 249–305.

[^2]: Edward L. Deci and Richard M. Ryan, self-determination theory — autonomy, competence, and relatedness as fundamental psychological needs; frustration of these needs produces genuine ill-being rather than mere dissatisfaction.