Pope Leo, “Artificial Intelligence,” and Catholic Social Teaching

By Jessica Sun
Pax Christi Young Adult Caucus

“Artificial intelligence” is poorly defined. It does not refer to a coherent set of technologies. Discussions about “AI” generally become more lucid when narrowed to the automation of specific tasks by performing statistical analyses of large datasets. “AI” is essentially a marketing term buoyed by Silicon Valley-funded “AI hype.” 

AGI, a subfield of AI known as “artificial general intelligence” or “AI singularity,” is another term that famously lacks a precise meaning. Bender and Hanna explain the myth of AGI in a publication from June. “Definitions range broadly in ways that primarily suit the economic arrangements of the individuals and organizations ostensibly trying to create it, or the cultural mystique of a set of adherents to a set of fringe ideologies. OpenAI’s charter defines AGI as ‘highly autonomous systems that outperform humans at most economically valuable work.’ Mark Zuckerberg has said that he does not have a ​​one-sentence, ‘pithy definition’ of the concept. Embodying the mysticism around the term, Ilya Sustkever, the former Chief AI Scientist at OpenAI, would lead chants of ‘Feel the AGI!’ around the office. In a leaked agreement, Microsoft and OpenAI crafted a much more well-defined metric: whether such a system could generate $100 billion in profit. In all these cases, the term is meant to evoke something with awesome power, much like the term ‘AI’ used to, before it became overexposed in marketing.”

In March, Brookings fellow Kyle Chan wrote on major differences in how the US is approaching AI versus China. “China’s AI companies are playing a different game. While they are also rushing to build world-class foundation models, the notion of ‘AGI’ as some abstract turning point in human history is less discussed. Instead, Chinese AI developers are racing along other axes of progress: efficiency, adoption, and physical integration, driven by both industry constraints and Beijing’s policy focus. Taken together, China’s approach is a fundamentally different bet on how AI will shape the future. … How can we ensure the US’s massive investment in AI data centers is sustainable and does not crowd out other areas of AI development or economic activity? Concerns have been raised that data center investments are distorting the broader economy, making up a large share of GDP growth, raising energy prices, and masking underlying problems like rising unemployment. If this issue is not addressed, the United States risks not only missing critical opportunities for AI innovation but also pulling the rest of the US economy into a giant, risky bet on AGI.”

A June 2025 report from Apple Machine Learning Research proved what had been clear to tech whistleblowers for years: Generative AI (GenAI) systems, which include systems that might lead to “AGI,” will never be capable of intelligence. GenAI is guessing, not thinking. GenAI is a subfield of artificial intelligence that uses generative models to generate text, images, videos, audio, software code or other forms of data. As artificial intelligence CEO Stephen Klein has summarized, large language models (LLMs), or “limited” language models like ChatGPT, are simply very expensive autocomplete. LLMs are arguably the world’s most expensive and energy-intensive plagiarism machine. “When you strip away clever prompting and pattern-matching tricks, large language models consistently fail at tasks that require multi-step reasoning or abstract logic. They appear to ‘solve’ problems only when the answers are memorized or embedded in the training data. Once you test them outside of that comfort zone, their performance collapses.” 

When is it safe to use ChatGPT? Source: UNESCO IESALC

A study this February on the language of artificial intelligence, “De-anthropomorphizing ‘AI’: From wishful mnemonics to accurate nomenclature,” explained that how we describe “AI” technology influences how it is perceived, deployed, and trusted. “It is the responsibility of journalists, companies, and scholars to characterize technology in ways that inform and empower their readers by using appropriate terminology and avoiding inflated claims,” reads the study. One type of inflated claim comes from using anthropomorphizing language to describe system functionality.” 

In a 2026 article from Tech Policy, the authors delve further into the problem with anthropomorphic descriptions, which risk masking important limitations of probabilistic automation systems that make them fundamentally different from human cognition. These are four of the examples they provided:

• Artificial intelligence > probabilistic automation
• Chatbot > conversation simulator
• Answer > output
• The agent did something > people used technology to do something

What a powerful difference already! Mystery AI Hype Theatre, a project of the Distributed AI Research Institute (DAIR), seeks to cut through the lies to explain the reality of what lies under the propaganda. DAIR is an independent organization conducting community-rooted research, a globally distributed group of academics, activists, and engineers who believe in technology that benefits everyone. Better Images of AI is a collaboration among various global academics, artists, diversity advocates, and nonprofit organizations. It aims to help create and curate a more representative and realistic visual language for AI systems, themes, applications, and impacts. The current dominant images of AI reinforce dangerous misconceptions, and at best limit the public understanding of the current use and workings of AI systems in their potential and implications. 

So, how did we get here? In a time of record wealth hoarding by the elite few, today’s new Gilded Age, tech oligarchs stand out in their influence and abuse of power. Thanks to the brilliant work of computer scientist Timnit Gebru and philosopher Émile P. Torres, we have a name for the mythologies driving tech power: TESCREAL, the cluster of ideologies quietly guiding tech’s most powerful decisions. Radical Futures created a TESCREAL quick guide that breaks it all down. 

Transhumanism

• Core belief: The human body is outdated. Technology should enhance or replace it.
• How it shows up: Brain chips. CRISPR edits. Longevity labs. Immortality startups.
• What it ignores: The wisdom of limitation. The value of impermanence. The politics of who gets “upgraded.”

Extropianism

• Core belief: Entropy is the enemy. With enough tech, we can defeat decay.
• How it shows up: Radical life extension. Energy-abundant futures. Post-death consciousness uploads.
• What it ignores: Grief. Cycles. The role of surrender. The sacredness of death.

Singularitarianism

• Core belief: A superintelligent AI will soon surpass humans. We must align with it.
• How it shows up: Alignment theory. AGI safety labs. Doom-and-salvation discourse.
• What it ignores: Current harm. Real workers. Non-sci-fi ethics. The political economy of AI now.

Cosmism

• Core belief: Earth is just a launch pad. Our destiny is among the stars.
• How it shows up: Mars colonization. Terraforming projects. Billionaire bunkers.
• What it ignores: Climate collapse here. Indigenous land. Stewardship. Repair.

Rationalism

• Core belief: Logic and reason are the only valid paths to truth. Emotion is a flaw.
• How it shows up: Hyper-quantified morality. Forum-based thought leadership. Cults of “intelligence.”
• What it ignores: Embodiment. Story. History. Feelings as knowledge.

Effective Altruism

• Core belief: To do the most good, we must maximize impact through evidence and logic.
• How it shows up: Spreadsheet philanthropy. Charity as optimization. Ethical career calculators.
• What it ignores: Context. Power. Community. The messiness of care.

Longtermism

• Core belief: The moral priority is to protect the far future, even over present lives.
• How it shows up: Billions of future people > current suffering. Ethics for hypothetical descendants.
• What it ignores: Urgency. Injustice. The people who are dying now.

As a friend remarked to me after he scrolled through the TESCREAL quick guide, “I’ve never seen such hatred of being human.” Catholic Social Teaching teaches us the importance of human dignity and the common good. Our Catholic faith calls us to address oppression, the communitarian role of the state, subsidiarity, social organization, social justice, and wealth distribution.

In November 2025, Pope Leo XIV spoke out on artificial intelligence. As Lever News Daily reported, he called on leaders to “cultivate moral discernment as a fundamental part of their work” and “develop systems that reflect justice, solidarity, and a genuine reverence for life.” Marc Andreessen — billionaire founder of tech investment firm Andreessen Horowitz and MAGA megadonor — responded by mocking the pope’s message, but received so much pushback that he deleted his post. Pope Leo is following in the footsteps of his namesake, Pope Leo XIII, who himself challenged the robber barons of the Gilded Age in the 19th century.

There is so much more to say about “artificial intelligence” and our Catholic faith, including and not limited to the destruction of our climate, worker exploitation in the Global South, colonized training data, novel mental health crises produced by the ChatGPT conversation simulator and the like, unaccountable power, and surveillance. Stop Killer Robots has sought to emphasize the danger of automated weapons in war. How does our theology transform when an AI weapon literally named “The Gospel” is being used by the State of Israel during its ongoing genocide of Gazan Palestinians, thanks to the blasphemy and horror that is Christian Zionism? Weapons technology is first tested on Palestinians, making it known as the Palestine Laboratory. 7amleh and Tech for Palestine are two organizations resisting tech for genocide on behalf of all of humanity. 

The Adalah Justice Project reminds us that from Palestine to Turtle Island, we must resist the attempted dismantling of human dignity powered by Big Tech. Take Skydio, a US company that makes AI-powered drones. In the first few weeks of Israel’s genocide, Skydio sent over a hundred drones to the Israeli military to be tested and honed on Palestinians. Since then, sales of “battle tested” Skydio drones have ballooned as police departments and security agencies across the US have signed new contracts with the company. Or we can look at Palantir, a company accused of providing the technology behind the Lebanon beeper attacks and under fire for its plans to lock Gaza down into zones of control. US immigrant rights and civil rights groups oppose Palantir for its use to track, target, and criminalize migrants and organizers.

Amid this darkness, there are also hopeful human efforts to nationalize AI in Norway, a Congressional campaign committing to zero AI-generated content, ethical AI startups such as Thaura AI, and growing alternative visions of AI, including the Small AI project. We need AI that is practical, secure, trustworthy, explainable, stable, and coherent under uncertainty. There is no path to that AI through pattern matching at scale. We can have a society with ethical tech, and there are many people sacrificing and working to create that reality right now. Last year, Denmark changed copyright law to ensure that everybody has the right to their own body, facial features, and voice. New York City just launched the Public Interest Technology (PIT) Crew. Also in the US, I am especially encouraged by the work of Corbin Trent and the New Consensus think tank in charting a path forward where AI might work for us all. 

New York became the first state to impose a needed data center moratorium. The US must update its energy grid to renewables and rebuild its crumbling infrastructure. It is time that the public has a say in discerning how, when, and if probabilistic automation should become such an integral part of society that we need any new data centers at all. The AI Data Center Moratorium Act of 2026 would institute an immediate federal moratorium on AI data centers until strong national safeguards are in place to ensure that: 1) AI is safe and effective—preventing executives in the AI industry from releasing harmful products that threaten the health and well-being of working families, our privacy and civil rights, and the future of humanity; 2) The economic gains of AI and robotics will benefit workers, not just the wealthy owners of Big Tech; and 3) AI does not increase electricity or utility prices, harm communities, or destroy the environment.

Empire of AI: Dreams and Nightmares in Sam Altman’s OpenAI (2025) by Karen Hao and The AI Con: How to Fight Big Tech’s Hype and Create the Future We Want (2025) by Emily M. Bender and Alex Hanna are two of the best books to shed more light on this essential issue of today’s world. 

As Pope Leo has explained, “The ability to access vast amounts of data and information should not be confused with the ability to derive meaning and value from it. … The latter requires a willingness to confront the mystery and core questions of our existence, even when these realities are often marginalized or ridiculed by the prevailing cultural and economic models.”

“The well-being of society,” he continued, depends on young people’s “ability to develop their talents and respond to the demands of the times and the needs of others, with generosity and freedom of mind.”

I was thankful for much of the intent in the pope’s first encyclical, Magnifica Humanitas, centering on artificial intelligence. I loved what Pope Leo said about protecting data workers and being discerning about the use of new technology that should be used to help, not harm, humanity. The Annotated Encyclical offers a reading of the encyclical by Hugging Face Society, which is an AI community on the journey to advance and democratize artificial intelligence through open source and open science. Despite the pope’s intent, I was struck by the hypocrisy of who the Vatican was consulting and centering.

When I was graduating college, the only then-“Big Five” tech company I was (barely) willing to sell my labor to was Microsoft. During my internship at Microsoft in 2018, I saw the company oppose employees protesting the sale of Microsoft technology to ICE. We do not have to accept a present where tech is used to hoard wealth, funneling power and resources into the hands of the few. 

Dr. Timnit Gebru is an embodied example of the heroism of a Big Tech whistleblower, one who put everything on the line to raise the alarm. Gebru is the founder and executive director of the Distributed AI and Research Institute, and co-founder of Black in AI. Google fired Gebru in December 2020 for raising issues of discrimination in the workplace, where she was serving as co-lead of the Ethical AI research team. 

Here is an excerpt from Gebru’s first thoughts on Magnifica Humanitas: “The Vatican could have told Anthropic to stop stealing data, exploiting labor, killing the environment, deceiving us with anthropomorphic designs, and lying about product ‘capabilities.’ Instead they partnered with them, like partnering with the Sackler family to discuss the harms of oxy. The Vatican could have partnered with the exploited data workers fighting for their rights, the people whose water is polluted fighting data centers, or the many other victims around the world.”

Dr. Milagros Miceli, Research Lead at the DAIR Institute, is a sociologist and computer scientist who works closely with data worker communities worldwide to interrogate labor conditions and power dynamics in the AI industry. In Miceli’s first thoughts, she adds: “It is also hard to imagine that Leo writes about the hidden labor powering AI without knowing that Anthropic outsources much of its data work under conditions very similar to those already exposed in relation to OpenAI and Meta. And the warning about trans- and posthumanism inevitably recalls what Timnit Gebru and Émile P. Torres have written about TESCREAL ideology, from which Anthropic is hardly exempt. Finally, it is difficult to read a text that echoes so many of the things many of us have been saying for at least a decade, while still covering them with a veil of inevitability. As if the only alternative were to imagine an AI aligned with the social doctrine of the Church, without leaving room to ask ourselves if we need AI at all.”

As James Baldwin well said, “I can’t believe what you say, because I see what you do.” I pray that the Vatican will do differently, and that each one of us will take an active role with technology in co-creating a future that protects and celebrates human dignity.

Take action

1) Tell the attorney general in your state to investigate your local law enforcement agency’s abuse of Flock surveillance cameras. Use this link to take action. (Click “Edit Subscription Preferences” to avoid opting into emails from all endorsing organizations.)

2) Tell lawmakers: Investigate big AI companies. Use this link to send a message to Congress.

3) New York has signed into law a moratorium on new hyperscale data centers. In Pennsylvania, Gov. Josh Shapiro has signed an executive order placing much needed limits on new data center projects — much better than nothing, but not as good as a real law that can out-last the whims of future governors.

Other states can do what these states have done. Use this link to email your state legislators and governor today.

---

Jessica Sun (she/they) is a second generation economic-settler-immigrant and Chinese-US American zillennial from northern Virginia. A Catholic convert after 10 years of identifying with varying philosophies as an agnostic, she finds a dear sense of belonging in spaces of zealous, Gospel-led advocacy and action for just peace and nonviolence. Christians for a Free Palestine, Pax Christi USA, and Progressive Asian American Christians are her organizing homes. A graduate of the Virginia Tech College of Engineering, she had worked as a software engineer by day, local activist by life. She recently began her postulancy with the Sisters of Providence of Saint Mary-of-the-Woods.

Featured image: A Rising Tide Lifts All Bots, Rose Willis & Kathryn Conrad / https://betterimagesofai.org / https://creativecommons.org/licenses/by/4.0/