Institutional religious responses to AI

Institutional religious responses to artificial intelligence include doctrinal notes, legal opinions, ethics statements, operational policies, pastoral guidance, interfaith declarations, and governance proposals issued by religious bodies. These documents commonly address human dignity, responsibility, truth, labor, privacy, warfare, authorship, religious authority, and the limits of delegating judgment to machines.

The documents do not have equal authority. A Vatican note issued by two dicasteries, a central-committee statement of the World Council of Churches, a responsum adopted by a rabbinical law committee, a fatwa from Egypt's Dar al-Ifta, an internal communications policy, and a multi-religious pledge perform different institutional functions. They should not be merged into a single religious position or treated as evidence that local clergy and communities have adopted the same practices.

The strongest detailed English-language sources in this article come from Christian, Jewish, and Islamic institutions. The multi-religious Rome Call broadens participation but does not substitute for tradition-specific documents from every signatory community.

Scope and source-authority rules

This article uses “institutional response” only for a document or policy attributable to a defined organization or authorized body. It distinguishes the authority of the issuer from the subject matter of the text.

Statements by individual clergy, academic theology, civil-society advocacy, and technology-company partnerships may influence institutional debate, but they are not treated as formal religious positions unless an authorized body adopts or issues them. A conference appearance or endorsement also differs from an operational policy: it can demonstrate participation without showing implementation.

Why AI became an institutional religious issue

Religious institutions encounter AI in two overlapping roles. First, they respond as moral and civic actors to technologies that affect employment, education, healthcare, war, surveillance, media, and political power. Second, they govern uses inside their own communities, including preaching, writing, communications, religious education, legal interpretation, pastoral care, and administrative work.

The Vatican's 2025 Antiqua et Nova treats AI as a broad social and anthropological issue, not merely a church-management tool. It examines the relationship between artificial and human intelligence and applies that distinction to human relationships, work, healthcare, education, misinformation, privacy, the environment, warfare, and relationship with God.^[1]

The WCC's 2023 central-committee statement arose from concern about rapid generative-AI development, commercial concentration, and weak regulation.^[2] By 2025, the WCC had also adopted an internal communications policy that recognizes uses such as translation, brainstorming, accessibility, meeting summaries, and research assistance while setting organizational conditions for responsible use.^[3] The movement from public ethics statement to operational policy illustrates the difference between normative concern and actual institutional adoption.

Human dignity and anthropology

Human dignity is the most consistent shared theme, but institutions ground it differently. Antiqua et Nova rejects reducing human intelligence to data processing or task performance. Its account emphasizes embodiment, relationships, moral responsibility, freedom, and orientation toward truth and the good. AI may extend human capacities, but it does not become a human person merely by performing cognitive tasks.^[1]

The Rabbinical Assembly's 2019 responsum approaches artificial agency through Jewish law. It concludes that current AI functions as a tool rather than a morally liable person: responsibility remains with the relevant human principal, even when designers, vendors, owners, and users may share responsibility. It also states that artificial agents may facilitate religious acts without becoming human or entering the covenantal status required for many obligations.^[5]

The Rome Call frames dignity in a shared public vocabulary rather than a complete religious anthropology. Its principles call for transparency, inclusion, accountability, impartiality, reliability, security, and privacy, and its signatories commit themselves to human-centered development.^[7] The overlap is significant, but agreement on governance principles does not imply agreement on the nature of the person, soul, consciousness, revelation, or salvation.

Truth, knowledge, and generated media

Institutional concerns about truth include fabricated content, misleading attribution, deepfakes, doctrinal error, and overconfidence in generated answers. Antiqua et Nova places responsibility on producers and users to verify generated material and avoid degrading, hateful, exploitative, or deceptive content.^[1]

The CCAR Responsa Committee addresses a narrower practical question: a rabbi using AI-generated poetry in worship. Responsum 5784.2 holds that the rabbi may describe his contribution but should not claim direct authorship or conceal the system's contribution. The committee's reasoning connects disclosure with honesty and responsibility: a generative system cannot be treated as an author in the same sense because it cannot answer for the work.^[4]

Dar al-Ifta similarly separates useful technology from authoritative religious judgment. Its 2025 fatwa states that AI is permissible in principle according to its purpose, but users may not rely on current AI applications to obtain fatwas. The institution grounds that restriction in the qualifications of a mufti, the need to evaluate sources and circumstances, and the inability of current systems to exercise the contextual legal judgment required for issuing a fatwa.^[6]

These are not identical rules. The CCAR text governs attribution of generated creative material; the Dar al-Ifta text governs reliance on AI for Islamic legal rulings. Both, however, preserve identifiable human responsibility rather than transferring authority to fluent output.

Labor and economic justice

Antiqua et Nova presents AI as capable of improving expertise, productivity, and creativity while warning that workplace systems can deskill workers, intensify surveillance, subordinate people to machine pace, and concentrate benefits among a small number of owners. Its stated priority is for technology to assist human work rather than treat workers as replaceable units.^[1]

The WCC's broader technology statements likewise connect AI with corporate concentration, unequal power, justice, and the social effects of rapid technological change.^[2] The Rome Call's principles of inclusion and accountability place economic and organizational decisions within a human-centered ethics framework, but the Call does not itself prescribe a detailed labor policy.^[7]

Institutional statements in this source set are primarily normative. They identify values and risks but provide little comparative evidence about how religious employers, schools, charities, publishers, or congregations are changing staffing, procurement, or compensation in response to AI.

Privacy and surveillance

Privacy concerns appear both as a general civil right and as a condition of religious freedom. Antiqua et Nova treats personal data as connected to relational and interior life, warns against surveillance that restricts freedom, and rejects reducing a person to a behavioral profile or social score. It calls for transparency, public accountability, and limits on those exercising surveillance power.^[1]

The Rome Call includes security and privacy among its six principles, while UNESCO's secular Recommendation places human rights, dignity, data governance, fairness, transparency, and human oversight at the center of AI policy.^[7][9] The overlap gives religious institutions a common policy vocabulary for public advocacy and procurement, but it does not answer tradition-specific questions about confession, pastoral records, membership data, or spiritually sensitive disclosures.

The current sources contain more general ethical principles than detailed denominational data-governance rules. A public endorsement of privacy does not establish that every affiliated institution has adopted technical controls, retention limits, or auditing procedures.

Warfare and autonomous systems

AI in warfare raises questions of lethal delegation, accountability, discrimination, escalation, and the moral status of machine decisions. Antiqua et Nova argues that weapons capable of identifying and striking targets without direct human intervention create grave ethical concern and places responsibility for the use of force on human agents and institutions.^[1]

The Rabbinical Assembly responsum considers autonomous transportation, medicine, warfare, and other life-or-death applications within a broader analysis of machine agency. Its conclusion that moral and legal responsibility remains with people is relevant to autonomous weapons, even though the responsum does not claim to settle every military-policy question.^[5]

The WCC's 2023 statement is directed primarily at unregulated development and concentrated power rather than a complete doctrine of autonomous weapons. The Rome Call's Hiroshima multi-religious event connected AI ethics with peace and brought leaders associated with eleven religions into a shared endorsement process.^[8] That event demonstrates interfaith participation; it is not equivalent to detailed and separately authorized teaching on war from each represented tradition.

Pastoral, clerical, and religious-legal uses

Operational policies tend to be narrower than public ethics documents. The WCC communications policy addresses uses in organizational communication, including translation, brainstorming, accessibility, summaries, and research assistance. Its authority concerns WCC communications, not all pastoral, liturgical, or doctrinal activity in member churches.^[3]

The CCAR responsum permits acknowledged human contribution to AI-assisted creative work while requiring disclosure of the system's role.^[4] Dar al-Ifta allows beneficial AI use in principle but prohibits reliance on open AI applications as substitutes for a qualified mufti.^[6] The Rabbinical Assembly text permits machines to facilitate actions while preserving human liability and denying current systems independent covenantal or moral status.^[5]

These examples support a recurring institutional pattern: assistance may be accepted under supervision, but authorship, legal interpretation, pastoral accountability, and ritual authority remain attached to authorized people and communities. The exact boundary is tradition-specific and cannot be inferred from technical capability alone.

Governance and accountability

Religious responses operate at several governance levels:

• Internal rules govern staff, communications, clergy, educators, or institutional publications.
• Religious-legal and doctrinal texts interpret established traditions in relation to new systems.
• Public ethics statements address governments, developers, companies, and civil society.
• Interfaith initiatives identify principles that multiple organizations can endorse without resolving theological differences.
• Partnerships with technology companies can support implementation and visibility but may also create source-proximity and accountability questions.

The Rome Call illustrates the partnership model. It began with the Pontifical Academy for Life, technology companies, the Food and Agriculture Organization, and an Italian government ministry; later events added Jewish, Muslim, Buddhist, Hindu, Zoroastrian, Bahá'í, and other religious signatories.^[7][8] The initiative establishes a shared commitment by named signatories, not independent proof that participating companies or institutions have fully implemented every principle.

UNESCO's Recommendation provides a useful external benchmark because it is an intergovernmental standard rather than a religious statement. Similar language about dignity, fairness, transparency, accountability, privacy, human oversight, and social justice can reveal areas of convergence, but religious documents may add claims about creation, covenant, vocation, conscience, pastoral care, or ultimate human purpose that the secular standard does not adjudicate.^[9]

Comparative coverage across traditions

The current English corpus does not support equally detailed sections for Hindu, Buddhist, Sikh, Jain, Bahá'í, Indigenous, new religious, and other institutional traditions. Participation in the Rome Call shows engagement by some leaders from several of these traditions, but a signature is not a substitute for a formal tradition-specific policy or legal text. The article therefore compares selected documented responses rather than claiming completeness.

Limits of the available evidence

Most sources are first-party institutional documents. They are the strongest evidence for what an issuing body formally stated, but weaker evidence for reception, compliance, local practice, or effectiveness. The UNESCO Recommendation supplies a high-authority external comparator, not independent verification that religious institutions have implemented their principles.

Authority also varies within traditions. A central institution, umbrella organization, committee, individual bishop, local congregation, academic theologian, advocacy group, and interfaith forum do not speak at the same level. English-language availability further skews the record toward institutions with well-developed international publishing systems.

Several documents are normative and prospective. They identify how AI should be governed, not how it is currently used. Claims about adoption require separate operational evidence such as policies, procurement rules, training programs, audits, disclosed deployments, or observed pastoral practice. Future revisions should add such evidence without presenting isolated examples as representative of an entire religion.

See also

• AI-generated worship and liturgy
• Religious chatbots and scripture interfaces
• Machine agency
• Machine-mediated religion

References

1. Dicastery for the Doctrine of the Faith and Dicastery for Culture and Education, Antiqua et Nova: Note on the Relationship Between Artificial Intelligence and Human Intelligence, 28 January 2025.
2. World Council of Churches Central Committee, “Statement on the Unregulated Development of Artificial Intelligence.” 27 June 2023.
3. World Council of Churches Central Committee, “Use of Artificial Intelligence in WCC Communications.” 24 June 2025.
4. Central Conference of American Rabbis Responsa Committee, Responsum 5784.2, “Artificial Intelligence and Authorship.”
5. Daniel Nevins, “Halakhic Responses to Artificial Intelligence and Autonomous Machines.” Committee on Jewish Law and Standards, approved 19 June 2019.
6. Egypt's Dar al-Ifta, “Using AI applications to obtain fatwas.” 2 December 2025.
7. Rome Call for AI Ethics.
8. “AI Ethics for Peace – Hiroshima, July 10th, 2024.” Rome Call, 12 July 2024.
9. UNESCO, Recommendation on the Ethics of Artificial Intelligence, adopted 23 November 2021.