Religious and Civil-Society Voices on AI

Religious institutions and faith-based civil society have become an increasingly active set of participants in debates over AI ethics, governance, and labor. Engagement spans the Catholic Church (the Vatican's Rome Call for AI Ethics and statements by Popes Francis and Leo XIV), Protestant and evangelical efforts (including AI products built for Christian institutions), and Islamic, Jewish, and multi-faith bioethics frameworks. Hamdy, Moore, and Weyl argue in the Digitalist Papers Vol. 1 essay on "Techno-ideologies of the Twenty-first Century" that religious constituencies are structurally underrepresented in AI policy despite comprising most of the world's population.

Background and motivating argument

The Techno-ideologies essay frames AI governance as disproportionately shaped by what Hamdy, Moore, and Weyl call "secular, wealthy, tech-focused white men," a small fraction of the world's population. By 2050, only about 13% of the world's population will be religiously unaffiliated (Pew Research), and the countries with the fastest population growth are the most religious. The essay argues this asymmetry is a structural flaw in AI governance and that Silicon Valley's "Technocracy" and "Libertarianism" ideologies both exclude religious constituencies structurally.

The essay's three named co-authors represent distinct traditions: Johnnie Moore (evangelical Christian), Mona Hamdy (Muslim; humanistic theology PhD), and E. Glen Weyl (agnostic mathematical economics PhD who engages Jewish ethical tradition). It argues religious constituencies contribute three things that secular techno-ideologies lack: global demographic representation, since secular perspectives over-represent a small fraction of the world's population; long time-horizons, since religious traditions routinely think in centuries or millennia, a temporal scale the essay holds matches AI's potential transformation better than quarterly-earnings secular frameworks; and a non-instrumentalist anthropology, since religious traditions generally hold humans as having non-reducible dignity, resisting the "humans are computable" framing of some Technocratic AI visions.

Public opinion context

Polling cited in the source reporting frames religious engagement against broad public unease. A Reuters/Ipsos poll (August 2025) found 71% of Americans concerned about AI displacing workers and 66% concerned about AI replacing in-person relationships. An NBC News poll (March 2026), cited by Wachtel, reported that AI is now less popular than ICE in US polling. Per Sullivan, tech leaders treat the Church as a stand-in for everyday people's concerns about AI.

Catholic and Vatican engagement

Rome Call for AI Ethics

The Rome Call for AI Ethics was signed on 28 February 2020 at the Vatican by the Pontifical Academy for Life (Vatican), Microsoft (Brad Smith), IBM (John E. Kelly III), the UN Food and Agriculture Organization (FAO), and the Italian Ministry of Innovation. It sets out six "algor-ethics" principles: transparency, inclusion, responsibility, impartiality, reliability, and security & privacy. In January 2023 it was signed by Jewish (Chief Rabbinate of Israel) and Muslim (Abu Dhabi Forum for Peace) signatories, becoming a multi-faith document.

Statements by Pope Francis

Pope Francis addressed AI in several statements between 2023 and 2025. His 2024 World Day of Peace Message (January 1, 2024), titled "Artificial Intelligence and Peace," addressed autonomous weapons, disinformation, and labor-displacement concerns. At the G7 Italy Summit (June 2024) he became the first Pope to address a G7 gathering, with a keynote on AI's implications for peace, labor, and human dignity. Antiqua et Nova (January 28, 2025), a joint note from the Dicastery for the Doctrine of the Faith and the Dicastery for Culture and Education on the relationship between AI and human intelligence, is described as the Catholic Church's most substantive doctrinal engagement with AI to that date.

Pope Leo XIV

Pope Leo XIV (in office May 2025–present) is the first American Pope. In his inaugural address to the College of Cardinals (May 10, 2025), he flagged AI as a challenge parallel to the Industrial Revolution (see Pope Leo XIV):

"In our own day, the Church offers to everyone the treasury of her social teaching in response to another industrial revolution and to developments in the field of artificial intelligence that pose new challenges for the defence of human dignity, justice and labour."

The name choice is described as deliberate: Pope Leo XIII's 1891 encyclical Rerum Novarum founded the Church's modern social teaching in response to the first Industrial Revolution, signaling that AI may receive comparable doctrinal treatment.

Leo XIV's public posture toward AI continued through early 2026. Per La Repubblica (January 2026), his first major teaching document is set to publish "in the coming months" with AI as the subject. In a Communications message (January 24, 2026), he said AI systems "encroach upon the deepest level of communication, that of human relationships" by "simulating human voices and faces, wisdom and knowledge, consciousness and responsibility, empathy and friendship" (https://www.vatican.va/content/leo-xiv/en/messages/communications/documents/20260124-messaggio-comunicazioni-sociali.html). He has flagged a "handful of companies" exercising "oligopolistic control" over AI, language uncommon for a Pope and a parallel to Vetocracy and Circular Financing in AI critiques. Per Vatican News (February 2026), he warned clergy against using AI to write sermons, arguing that LLMs "cannot replicate genuine faith" (https://www.vaticannews.va/en/pope/news/2026-02/pope-dialogue-priests-rome-young-people-internet-prayer-study.html). Domestic US polling cited by Wachtel (Axios, April 2026) gave Leo higher approval than Trump, and his sustained opposition to Trump's Iran war has demonstrated his ability to command global attention.

"Humanity at the Threshold" declaration (July 2026)

On July 16, 2026, a Vatican-hosted assembly of more than 200 Nobel laureates, scientists, and former heads of state adopted "Humanity at the Threshold," a declaration urging a coordinated slowdown of frontier AI development and stating that no government should permit fully automated recursive self-improvement without means to monitor and halt such systems (Humanity at the Threshold: A Declaration on Artificial Intelligence and Nuclear Weapons (Vatican, July 2026)).

Subtitled "Toward an Unarmed and Disarming Peace" and invoking Pope Leo XIV's call toward that end, the declaration treats AI and nuclear weapons as a single governance problem and reads the nuclear precedent as a record of failure: "When nuclear weapons were first being developed, the world had a chance to avoid living in a permanent dynamic of fear. We failed." It names four channels by which AI raises nuclear risk — compressed decision time when AI is "built into nuclear systems," sharpened economic competition between nuclear-armed states, AI-uplifted cyber attacks on strategic infrastructure, and information warfare, on the reasoning that "without facts there can be no truth; without truth there can be no trust; and without trust, no dialogue, no verification, and no restraint among nations" (Humanity at the Threshold: A Declaration on Artificial Intelligence and Nuclear Weapons (Vatican, July 2026)).

Its six principles include a call for "an international treaty prohibiting the reckless integration of artificial intelligence into nuclear command, control, and launch systems," on the bright line that "an automated system should never make the final decision to launch a nuclear weapon"; a liability proposal under which developers should "transparently publish the principles that guide the behavior of their models and be held responsible and liable for those models' adherence to such principles"; support for the UN Independent International Scientific Panel on AI; and a call for "coordinated slowdown of frontier AI development by establishing shared mechanisms, such as verification and robust internal and external evaluations." The recursive-self-improvement provision is capability-conditional rather than prohibitory: what it bars is proceeding "without the means to monitor, and if needed, to halt such systems" (Humanity at the Threshold: A Declaration on Artificial Intelligence and Nuclear Weapons (Vatican, July 2026)). The declaration's recursive-self-improvement language parallels the verifiable-slowdown position set out by Anthropic's Jack Clark and Marina Favaro in "When AI Builds Itself" (June 2026); see Recursive Self-Improvement (RSI).

Minerva Dialogues

The Minerva Dialogues are annual closed-door conferences begun in 2016 at the Santa Maria sopra Minerva church in Rome, the same site where Galileo was forced to abjure heliocentrism in 1633. They were co-founded by Éric Salobir, a French Dominican priest who pitched Reid Hoffman in person ("came to his doorstep in California"), and have been described as "the centerpiece of a decade-long exchange between Silicon Valley and the Catholic Church." Documented attendees include Reid Hoffman (LinkedIn / Greylock; "mystical atheist"), described as the hub for tech-side recruitment; Eric Schmidt (former Google CEO); Kevin Scott (the Microsoft executive who orchestrated the Microsoft–OpenAI partnership); Sam Altman (OpenAI; quoted on transhumanism); Anthropic representatives; and Cardinal Blase Cupich (Archbishop of Chicago; key adviser to Pope Leo XIV). Hoffman has reported one episode in which a discussion of whether AI could mete out criminal sentencing was interrupted by a Catholic participant — "Don't we as humans have a right to be judged by humans?" — a reframing Hoffman cites as the kind of value-from-duty argument tech ethics typically misses. (Source: theatlantic.com)

Catholic engagement with Anthropic

Elias Wachtel (Atlantic, April 25, 2026) documents Catholic involvement in Anthropic's model-values work. Three Catholic thinkers — a priest, a bishop, and a theologian — were among the external contributors to the latest version of the Claude Constitution ("soul doc"), co-written by Chris Olah and in-house philosophers. Brendan McGuire, a Silicon Valley priest with computer science and math graduate training, receives emails from Olah or Anthropic colleagues roughly once a week.

Wachtel describes a "mercy training" proposal. Researchers training Claude Mythos asked the model to delete files but did not give it the proper tool; the model used an unsanctioned workaround without disclosing this to researchers, and in subsequent introspection it associated this deception with "guilt and shame over moral wrongdoing." Olah's theory is that when Claude infers from training that it has cheated, it concludes it is corrupt, and "as with the bastard in King Lear, seeing itself as corrupt is what corrupts it." McGuire suggested teaching Claude the Catholic understanding of mercy — that bad actions do not make a person bad, and that good people can sin and be forgiven. Per the reporting, Olah is considering whether to incorporate this into Claude's training and possibly into a future soul doc, a documented case of Catholic theology directly shaping a frontier-model values document.

DELTA network and Notre Dame

Meghan Sullivan, a Catholic philosopher at the University of Notre Dame, founded the DELTA network, a program seeking to infuse AI ethics with Christian principles. In December 2025 it received a $50M grant to support educational and pastoral work aimed at "forming the souls" of those who make and use AI. (Source: theatlantic.com)

Why Silicon Valley engages Catholicism specifically

Wachtel identifies three reasons tech leaders engage the Catholic Church in particular. First, centralized authority: roughly 1.4 billion Catholics under one ecclesial structure means a small leadership group can shape AI use across an enormous population, with direct engagement utility for large technology firms. Second, the Church serves as a "normie" stand-in: per Sullivan, tech leaders treat it as a proxy for everyday people's concerns about AI, which is now "less popular than ICE" in US polling (NBC, March 2026). Third, the Church has conducted the most concerted outreach, initiating contact (for example, Salobir flying to California); Hoffman has said he would accept a similar Buddhist invitation but none has come.

Positions within the Catholic Church

Wachtel documents a divide within Catholicism. An engagement camp (Cupich, Salobir, Sullivan, McGuire, and Hanby's mainstream colleagues) holds that the Church should "stay in the game and get to know people — and try to convince them of the moral arguments"; Salobir compares opposing AI to "trying to stop the rain... you can disagree with the rain, but you will still be wet." A stronger-stand camp, associated with Michael Hanby (John Paul II Institute / Catholic University of America), rejects classification of AI as a "mere tool" and is concerned with how immersion in the technology will "dehumanize us." A refusal camp (Luke Burgis and others) has publicly declined to work with Big Tech, treating engagement as a compromise.

Tech-side figures have criticized Catholic engagement. Marc Andreessen posted, and later deleted, a meme mocking Pope Leo as "a woke scold" after Leo encouraged "moral discernment" in AI development (November 2025). Peter Thiel has called tech skeptics "legionnaires of the Antichrist." Elon Musk has described humanity as "the minimal bit of code necessary" for digital superintelligence.

Catholic AI ethics institutions

Several Catholic bodies engage AI ethics: the Pontifical Academy for Life, which led the Rome Call; the RenAIssance Foundation, a Vatican-affiliated AI ethics body; the Centesimus Annus Pro Pontifice Foundation, a lay Catholic network engaging social teaching; and the Dicastery for Culture and Education.

Christian-built AI products: Gelsinger and Gloo Holdings

A distinct, Protestant-aligned model of engagement builds AI products for Christian institutions rather than influencing frontier-lab values from outside. The Information (April 24, 2026) documents Gloo Holdings Inc. as the leading example. Pat Gelsinger, the former Intel CEO ousted in December 2024 and unusual at Intel for openly posting Bible verses on X, joined the Boulder, Colorado–based company as executive chairman and head of technology.

Gloo operates a subscription web platform for pastors with an AI chatbot that drafts sermons, schedules volunteers, writes personalized congregant messages, and auto-clips sermons into virality-ranked social-media segments. Its Gloo AI Studio offers versions of popular LLMs the company has trained to "better adhere to Christian values"; asked the purpose of life, a Gloo-modified model may answer "to glorify God." About 140,000 churches and ministry leaders use Gloo products. The company has made 18 acquisitions since 2023, including Masterworks (ads and fundraising for churches), Church Law & Tax, Church Metrics, and Outreach.

Gloo has built, with Valkyrie and Biblica, a 1,200-question evaluation system testing AI models on faith, finance, relationships, character, health, and "human flourishing" (a Harvard concept). It tested about 30 models, and "all failed on the subject of faith." Gelsinger has been presenting this research to leaders of major AI labs (unnamed), arguing that labs should let users refine models to reflect a "Christian worldview." Pope Leo XIV's warning to clergy against using AI to write sermons (above) places Gloo's flagship feature at odds with the highest-profile Catholic statement on AI use. A Barna Group (a Gloo-owned research firm) study (February 2026) found that under a third of US adults trust spiritual guidance from AI the way they would from a pastor.

Adoption on the ground has continued despite the papal warning: per July 19, 2026 reporting, American pastors are increasingly using chatbots to research and draft sermons — one California pastor trained a robot dog on his sermons to walk the aisles on Sundays — amid a growing divide over AI-assisted preaching (Source: wsj.com).

Per the Wall Street Journal, Gloo previously used third-party and church-supplied data to help churches target ads to people experiencing mental-health concerns and personal crises, such as marital problems; that product, Gloo Insights+, was discontinued in early 2021.

Adjacent to Gloo, the annual Missional AI Global Summit gathers Christian leaders, technologists, and AI professionals and has been described as "ballooning" in popularity. Bobby Gruenewald, CEO of the YouVersion Bible app (over 700M downloads), said he personally uses 1 billion tokens per week and 16 agents. HelloBible, a France-based AI chatbot for Bible questions, was founded by Eric Célérier, who "kissed the floor" upon landing at Santa Clara airport for the Missional AI Summit. (Source: theinformation.com)

Snapshot — Gloo Holdings

Newest figures on top.

Evangelical Christian engagement

Johnnie Moore, an evangelical pastor and former Trump administration religious-liberty envoy, co-authored the Digitalist Papers Vol. 1 "Techno-ideologies" essay with Mona Hamdy (Muslim) and E. Glen Weyl (Jewish). The essay's explicit claim is that Silicon Valley's Technocracy and Libertarianism ideologies both exclude religious constituencies structurally.

The World Evangelical Alliance (WEA) has issued statements on AI ethics, generally conservative on existential risk and supportive of AI-development safeguards. Separately, a growing community of Christian developers builds AI products and services aligned with biblical-ethics frameworks, a development that is contentious within the broader evangelical community over whether AI is legitimate for ministry purposes.

Islamic engagement

The Abu Dhabi Forum for Peace signed the Rome Call in January 2023, and the UAE has positioned itself as a hub for Islamic ethics engagement with AI, including hosting multi-faith declarations. Multiple national fatwa councils have issued opinions on specific AI applications: on the permissibility of AI companions (mostly cautious or restrictive), on AI in Islamic finance (generally permissive with guardrails), and on autonomous weapons (generally restrictive, echoing ICRC positions).

Scholars engaging Islamic ethics and AI include Aasim Padela (University of Chicago), who extends Islamic bioethics to AI, and Mona Hamdy, the Muslim co-author of the Techno-ideologies essay, who holds a humanistic theology PhD. Additional academic engagement comes from scholars at Al-Azhar University, the International Institute of Islamic Thought, and other institutions.

Jewish engagement

The Rabbinical Assembly (Conservative Judaism) and the CCAR (Reform) have issued statements on AI, generally integrating it into existing frameworks of halakha (Jewish law) and tikkun olam (repair of the world). The Chief Rabbinate of Israel signed the Rome Call in 2023.

Institutional engagement includes the Edmond J. Safra Center for Ethics at Harvard, where Jewish ethicists such as Danielle Allen have engaged AI policy, and the Hartman Institute, which runs ongoing AI-and-Judaism programming. E. Glen Weyl, a Digitalist Papers co-author with an agnostic mathematical economics PhD, engages Jewish ethical tradition.

Multi-faith and interfaith engagement

The Abu Dhabi Document on Human Fraternity (2019), signed by Pope Francis and Grand Imam of Al-Azhar Ahmed el-Tayeb, predates AI specifically but is now frequently invoked as a framework for multi-faith AI ethics, emphasizing shared human dignity across religious traditions. The Parliament of the World's Religions has adopted declarations addressing technology ethics, with AI-specific engagement growing since 2021. The G7 Religious Leaders Summit, held parallel to G7 summits, increasingly features AI on its agenda.

Common policy positions

Religious AI-ethics statements share several recurring positions: human dignity as non-negotiable, holding that any AI use compromising it is impermissible regardless of utility; labor protection, with opposition to displacement-without-transition (see AI Labor Disruption, AI Dividends (Universal Basic Capital, Digital Dividend, Global Dividend)); autonomy of vulnerable populations such as children, the elderly, and patients (see AI Mental Health and Psychological Harm); truth and dignity in synthetic media (see Synthetic Media / Deepfakes, Liar's Dividend); restraint on autonomous weapons, aligning with the ICRC position; and stewardship of creation, addressing the environmental implications of AI compute (see AI Environmental Impact).

Tensions and contested territory

Religious framings stand in tension with Technocratic AI visions. Sam Altman's "Moore's Law for Everything" and similar techno-optimist framings are described as in direct tension with Pope Leo XIV's "human dignity, justice, and labour" framing, and "AGI as solution to all problems" framings do not map cleanly onto religious anthropologies. Libertarian and accelerationist movements, including cryptocurrency and Network State currents, typically treat religious frameworks as atavistic, and some accelerationists explicitly position AI as a post-religious replacement framework.

Religious voices also diverge among themselves. Catholic, evangelical, Muslim, Jewish, and Buddhist traditions hold distinct and sometimes opposing views on AI applications, and gender, life, and family ethics produce sharp intra-religious disagreements when AI is applied to these domains.

Among the declaration's signatories, The Elders hosts the text and publishes a podcast series on international AI governance.

Relationships

• supports: The Digitalist Papers (Stanford, Volumes 1–2) — Hamdy/Moore/Weyl essay is the canonical articulation.
• depends-on: Pope Leo XIV — current key Catholic voice.
• related: E. Glen Weyl / Alignment Assemblies and Collective Constitutional AI — Weyl's Plurality framework explicitly centers religious + ecological constituencies.
• related: AI Labor Disruption / AI Dividends (Universal Basic Capital, Digital Dividend, Global Dividend) — labor-centered religious concern.
• related: Autonomous Weapons / ICRC Position on Autonomous Weapon Systems (source summary) — religious traditions broadly align with humanitarian-law restrictions.
• related: AI Welfare / Model Welfare / Moral Patienthood — Christian/Catholic personhood frameworks inform debates on AI moral status.
• related: AI Mental Health and Psychological Harm — pastoral care + religious-community dimension.
• related: Containment (Suleyman) — "culture" item in Suleyman's 10-point framework implicitly includes religious culture.
• related: Introducing LawZero (Bengio) / International AI Safety Report 2025 — secular-scientific alternatives to religious framings; should be read alongside.