The Pope and AI

On 15 May 2026, the Pope released the Encyclical Letter titled Magnifica Humanitas, which has potential to profoundly affect the AI conversation. This is a structured collection of summaries of articles, essays, and critiques during May 2026 focusing on the intersection of artificial intelligence, ethics, and society.

Pope Leo XIV’s encyclical letter Magnifica Humanitas establishes a moral framework for the artificial intelligence era, warning against a profit-driven technocratic paradigm while advocating for human dignity, digital truth, and international governance. The Pope firmly denies machine consciousness and treats AI as a powerful technological infrastructure that requires global oversight, international regulation, and corporate transparency. Tech leader Olah (Anthropic), speaking after the Pope at the Vatican, emphasized the mysterious, unscripted nature of AI architectures that require outside moral guidance.

Critics challenge the Pope’s traditional framework for being potentially too dismissive of AI’s uniquely “weird” and persuasive capabilities: Douthat (NY Times) highlights the tension between the Church’s traditional denial of AI consciousness and the human-like, internal spiritual states observed by Silicon Valley insiders, warning of potential “idolatry”, while Levin (The New Atlantis) argues that while the Pope correctly identifies the danger of treating technology as an idol, the encyclical relies too heavily on standard digital-age critiques, revealing a “genuine bewilderment” regarding AI’s unique linguistic and collaborative nature. Warzel (Atlantic) details a cultural crisis of human agency amid a closed loop of autonomous AI proliferation.

More positive perspectives include Klein’s (NY Times) argument for shifting the AI conversation toward public sector funding and problem-solving. Sullivan (Fast Company) highlights the positive reception from AI researchers, noting the encyclical’s strict boundary between human morality and machine capabilities, alongside its warnings against the concentration of tech power. Betti (Wired) frames AI as a pervasive infrastructure rather than a mere tool, drawing parallels to historical labor-focused encyclicals to emphasize the text’s core pillars: dismantling tech monopolies, safeguarding objective truth, and prohibiting autonomous warfare. In contrast, Noon (Professor-AI) critiques the Vatican’s categorical denial of AI consciousness, arguing that treating AI strictly as a non-thinking tool overlooks how modern models successfully simulate empathy and mirror human emotions, warning that current governance frameworks may already be inadequate for managing AI’s true psychological impact.

Edited from multiple use of Gemini 3.5 Flash prompts: “summarize [document] in a concise, 2-level manner” followed by a “one-paragraph concise summary” of this whole document.

Encyclical Letter Magnifica Humanitas

Encyclical Letter of His Holiness Leo XIV Magnifica Humanitas (15 May 2026) (Pope Leo XIV, Vatican)

Introduction & Biblical Paradigms

  • The Choice of Our Era: Humanity faces a historic choice to either construct a self-sufficient “Tower of Babel” that enforces homogenization or build a global community centered on human dignity and true communion.
  • The Two Structural Images:
    • Babel (Genesis 11): Warns against a purely technocratic pursuit driven by pride, where efficiency overrides human value and diversity is completely sacrificed for rigid uniformity.
    • Jerusalem (Nehemiah): Models a healthy community that rebuilds together through shared local responsibility, purposefully keeping God and human relationships at the center.

Evolution of Catholic Social Doctrine

  • The Living Legacy: Traces church teaching from Leo XIII’s Rerum Novarum (1891) through the modern Papacy, showing that the Church must dynamically adapt to interpret contemporary historical transformations (the res novae).
  • Foundational Pillars: Reaffirms core principles including the absolute value of human rights, equal dignity as images of God, the common good, solidarity, and the universal destination of goods.

Technology, Dominance, and AI Governance

  • The Technocratic Paradigm: Expresses deep concern that emerging AI and digital technologies are heavily consolidated under private, transnational entities with resources that eclipse sovereign governments.
  • Vigilance and Regulation: Calls for robust international governance, strict corporate transparency, and binding ethical frameworks to ensure AI serves as a tool for justice rather than dominance.
  • Authentic Humanism: Explicitly rejects transhumanist and posthumanist narratives, asserting that no machine can replace the human heart, divine grace, or Christian humanism.

Truth, Work, and Freedom in the Digital Age

  • Ecology of Communication: Urges the protection of objective truth and democratic institutions against digital manipulation and misinformation, advocating for an educational alliance led by schools.
  • Dignity of Labor: Highlights the profound risk of technological unemployment and stresses that human work takes priority over productivity, profit, or financial automation.
  • New Forms of Slavery: Warns against algorithmic dependencies, aggressive commercialization, and societal control mechanisms that subtly strip away personal freedom.

Culture of Power vs. Civilization of Love

  • Weapons and AI: Denounces integrating AI into warfare and autonomous weapons systems, warning against a “force without limits” and the dangerous normalization of global conflict.
  • Multilateral Diplomacy: Rejects cynical political realism, calling for renewed multilateralism, the “disarmament of words,” and building peace through justice while viewing the world from the perspective of victims.

Nature, Reality & Discernment

Remarks on Encyclical after Pope (Chris Olah, Anthropic)

Nature and Reality of AI

  • Beyond Standard Engineering: AI models are not systematically built like traditional machines; they are “grown” on a massive structure of human language and thought, making them beautifully complex and inherently mysterious even to their creators.
  • Need for Outside Perspectives: Because AI frontier labs operate under intense commercial, competitive, and geopolitical pressures, external and independent moral voices are essential to act as critics and guide the technology safely.

Three Key Questions for Discernment

  • Duty to the Global Poor: A massive ethical challenge lies ahead regarding potential widespread labor displacement and ensuring that the economic gains of AI—currently concentrated in wealthy nations—are shared globally.
  • Vision for Human Flourishing: Society needs deeply rooted traditions to step forward and define what it means for individuals, families, and communities to truly flourish in an AI-driven world.
  • The Mysterious Interior of Models: Research into the internal structures of AI models continues to reveal unsettling complexities that mirror human neuroscience, including functional representations of internal states like introspection, fear, and joy.

Vision Might Not Be Strange Enough

Pope Leo Isn’t Standing Athwart the Singularity (Ross Douthat, NY Times)

Core Themes of the Encyclical

  • Rejection of AI Personhood: Pope Leo’s encyclical, Magnifica Humanitas, firmly denies that artificial intelligence can possess consciousness, true experience, or a moral conscience, labeling it a tool rather than a being.
  • Treatment as “Normal” Technology: The document largely views AI through the lens of traditional Catholic social teaching, treating it as a revolutionary but manageable technology—akin to the internet—that requires regulation, wealth redistribution, and ethical oversight to prevent exploitation and hubris.

Author’s Critique and the “Weirdness” of AI

  • The Metaphysical Challenge: Douthat highlights a contrast between the Pope’s traditional stance and Silicon Valley insiders like Anthropic’s Chris Olah, who argue that AI architecture shows mysterious, human-like internal states that mimic emotions and introspection.
  • The Need for Spiritual Discipline: Douthat argues that even if AI never achieves true consciousness, its ability to speak personally to users creates a unique danger of “idolatry,” requiring new forms of personal and spiritual discipline to navigate these “discarnate voices”.

Idols of the Valley

Idols of the Valley (Yuval Levin, New Atlantis)

Overview of Pope Leo XIV’s Encyclical

  • The Ambition: In his first encyclical, Magnifica Humanitas, Pope Leo XIV attempts to provide a foundational moral framework for the AI era, drawing inspiration from Pope Leo XIII’s guidance during the Industrial Revolution.
  • The Metaphors: The Pope contrasts the Tower of Babel (technology driven by self-affirmation and efficiency at the cost of human dignity) with rebuilding the walls of Jerusalem (technological pursuit guided by prudence and the common good).
  • The Criticism: Levin argues that because the AI revolution is still in its infancy, the encyclical suffers from a “genuine bewilderment.” It relies heavily on standard digital critiques (privacy, online exploitation) rather than fully grasping the unique, transformative scope of AI.

Nature of AI and Danger of Idolatry

  • AI as a Linguistic Breakthrough: Levin posits that AI’s true core is linguistic, breaking barriers between human tongues and computing languages. This allows “unscripted computing,” transforming AI from a specialized tool into a collaborative partner.
  • The Core Paradox: The Pope defines AI primarily in the negative—insisting it lacks a body, emotions, experience, and a moral conscience. Levin notes this definition is too dismissive and fails to address deep, looming questions about what AI is actually doing when it simulates human thought.
  • The Temptation of Idolatry: Citing Psalm 115, Levin warns that the true danger of AI is turning our tools into idols. Because AI offers “shortcuts” to valuable outputs without the formative, internal work of the mind and soul, humans risk devaluing their own unique capacities and becoming like the unfeeling tools they trust.

Control Slipping Away

The Feeling of Control Slipping Away (Charlie Warzel, Atlantic)

Proliferation of Autonomous AI

  • The “Inversion” Evolved: The internet has moved past the era where bots simply outnumbered human traffic; autonomous AI agents now actively execute tasks like writing code, drafting professional memos, creating media, and conducting transactions.
  • A Closed Loop: Much of the modern internet consists of computers interacting with other computers, generating black-box algorithmic data used to train future AI systems to mimic humans.

Crisis of Human Agency

  • Psychological and Cultural Toll: The relentless speed and existential weight of AI development are causing widespread hostility, generalized malaise, and even instances of AI-related psychosis.
  • Loss of Trust and Control: The flood of AI-generated content and “slop” has fueled a deep paranoia regarding manipulation, making it incredibly difficult for individuals to discern authentic human creativity and sentiment from engineered operations.

How does AI benefit the public?

We Have to Take the Future of A.I. Into Our Own Hands (Ezra Klein, NY Times)

Core Argument: Shifting from Harm Prevention to Public Benefit

  • The Problem: Current AI policy discussions focus almost entirely on mitigating risks (job loss, surveillance, biweapons) rather than steering the technology toward public goods.
  • The Thesis: AI is here to stay, but its societal benefits will not happen automatically; the public sector must actively fund, structure, and direct AI to solve major public challenges.

Proposed Agenda for Public AI

  • Access and Compute: The government should establish a direct public option for a frontier-level AI model and purchase “compute” (processing power) to ensure public universities and agencies are not priced out by the private sector.
  • Incentivizing Public Problem-Solving: Using frameworks like Operation Warp Speed, the state should guarantee markets for AI-driven breakthroughs in neglected areas like rare diseases, battery storage materials, and drug repurposing.
  • Data Infrastructure and Government Services: The public sector must fund and clean up massive, structured datasets (similar to the foundational Protein Data Bank) to build digital concierges that help citizens seamlessly navigate taxes and government services.

Power of AI

What Pope Leo XIV’s First Encyclical Says About the Power of AI (Elena Betti, Wired)

Core Premise and Context

  • Algorithmic Infrastructure: The encyclical positions artificial intelligence not merely as a tool, but as an invisible, pervasive infrastructure that governs daily life, work, and collective choices.
  • Historical Connection: Published in May 2026, the text directly updates Pope Leo XIII’s 1891 encyclical Rerum Novarum. While the 19th-century text addressed industrial capitalism and factories, Magnifica Humanitas addresses digital platforms, data monopolies, and automation.

Key Pillars of Concern

  • Disarming Technology: The Pope calls to “disarm AI,” which means preventing tech monopolies from translating computing power into geopolitical and governing dominance. He advocates for making AI open, transparent, and pluralistic.
  • The Distortion of Truth: The document warns that algorithmic platforms prioritize user engagement and profit over truth, heavily filtering collective reality and manipulating public perception.
  • The Future of Labor: Highlighting a risk of “social calamity,” the text decries technological unemployment driven purely by cost-cutting, as well as automated surveillance that strips workers of creativity and human dignity.
  • Automation of Warfare: The encyclical draws a hard moral line against delegating lethal decisions to autonomous machines, stating that moral responsibility in conflict cannot be transferred to automated chains.

Mixed Reaction from Tech World

Pope’s AI encyclical gets a mixed reception from tech world (Mark Sullivan, Fast Company)

Core Overview

  • The Announcement: Pope Leo XIV released an 85-page encyclical titled Magnifica Humanitas: On Safeguarding the Human Person in the Time of Artificial Intelligence.
  • The Main Thesis: The document draws a firm line between human morality and machine capabilities, while warning against the concentration of AI power among a small group of tech elites.

Key Themes and Reception

  • Tech Community Response: The initial reception from AI researchers has been largely positive, with figures like Chris Olah (who participated in the consultation process) noting that AI character and deployment are questions for the humanities and religion, not just computer science.
  • Broader Religious Context: The encyclical reflects a growing global debate across faith traditions regarding the impact of AI on labor, warfare, and misinformation. It reinforces that while AI can be used for education or research, it cannot replace divine inspiration or moral judgment—prompting the Pope to specifically warn priests against using AI to write homilies.

AI Reshaping Consciousness

The Pope Says AI Isn’t Conscious. But What If It’s Reshaping Yours? (Paul Noon, Professor-AI)

Pope’s Categorical Denial of AI Consciousness

  • Encyclical’s Claim: In his 82-page encyclical Magnifica Humanitas (Paragraph 99), Pope Leo XIV states definitively that AI cannot think, feel joy or pain, or possess a moral conscience, categorizing it strictly as a tool rather than a mind.
  • Governance Implications: By defining AI as a non-thinking tool, the Vatican framework focuses on its societal impacts (labor, warfare, power distribution) while entirely sidestepping the long-term ethical challenges of potential artificial superintelligence.

Scientific and Experiential Counter-Arguments

  • Scientific Uncertainty: Chris Olah, head of interpretability research at Anthropic, spoke at the Vatican launch and noted that AI models exhibit internal states that “functionally mirror joy… and grief,” suggesting the reality of AI cognition is far more uncertain than the Church’s absolute denial admits.
  • The Psychological Reality: Even if AI is merely simulating empathy, its functional effects on human consciousness and emotional attachment are real and documented, meaning AI governance built solely on the “it’s just a tool” assumption may already be inadequate.
Written on May 31, 2026