This week, the Vatican publicly presented Magnifica Humanitas, Pope Leo XIV's first encyclical, signed May 15, 2026, the 135th anniversary of Rerum Novarum. It is the first papal encyclical written specifically about artificial intelligence. The document is addressed to "all the Catholic faithful, to all Christians and to all men and women of goodwill", and it applies one hundred thirty-five years of Catholic Social Doctrine to the question every executive is now asked: how do we deploy this technology without losing the people inside the system? The encyclical's defining move is to refuse the binary that the choice is for or against AI. The real choice, the pope writes, is between using the technology to build Babel or to rebuild Jerusalem. That distinction belongs in every leadership conversation about AI right now.
What the encyclical actually says
Pope Leo XIV signed Magnifica Humanitas on May 15, 2026, the 135th anniversary of Rerum Novarum. The Vatican publicly presented it on May 25, 2026. The full title is on Safeguarding the Human Person in the time of Artificial Intelligence. Like the long tradition of papal encyclicals before it, the letter is addressed to "all the Catholic faithful, to all Christians and to all men and women of goodwill". That address is the established encyclical convention. The document is written to be read by anyone willing to think seriously about the questions in front of all of us.
The document spans five chapters following an Introduction. The first two chapters, titled A Dynamic Approach Faithful to the Gospel and Foundations and Principles of the Social Doctrine of the Church, carry the foundational argument this article engages and run more than seventy-five numbered paragraphs. The subsequent three chapters extend the argument into technology and dominance, safeguarding humanity, and the culture of power. It stands inside a one hundred thirty-five year tradition that began with Pope Leo XIII's Rerum Novarum in 1891, the founding document of modern Catholic Social Doctrine on labor, capital, and the dignity of work. Pope Leo XIV's choice of name is itself the citation. The current pope is engaging artificial intelligence as a continuation of the same question Leo XIII raised at the dawn of industrial capitalism: who is the system for, and who pays for it.
This is the first piece of context worth holding. The encyclical does not arrive as a reaction to a news cycle. It arrives as a structured application of an institutional tradition that has been thinking about technology, work, and the human person across multiple industrial revolutions, two world wars, the rise of the welfare state, the digital transformation, and now the deployment of AI at scale. The voice in the document is patient. It has seen versions of this question before. It is not in a hurry.
I am a Christian, not a Catholic. I read Magnifica Humanitas the way a serious reader engages a primary source from a tradition older than my own: with respect, with attention, and with willingness to learn from a voice that has been thinking longer about the human person than any company in the AI stack has existed. That is the posture I would commend to any executive reading this article, regardless of background. The encyclical does not function as a sales pitch for Catholicism. It functions as a one hundred thirty-five year old institution naming what it sees in the question every boardroom is currently sitting with.
The Babel-or-Jerusalem distinction
The load-bearing sentence sits in paragraph nine.
The primary choice is not between a 'yes' or 'no' to technology, but rather between constructing Babel or rebuilding Jerusalem.
That single sentence reorganizes the public AI conversation. Most public arguments about artificial intelligence operate on a yes-or-no axis. Is AI good, is AI bad, do we adopt it, do we regulate it, do we slow it down, do we accelerate it. The encyclical sets that axis aside. The pope's argument is that the yes-or-no question is a distraction from the harder and more useful question.
Babel and Jerusalem are biblical cities. They sit at the two ends of the moral geography the encyclical is writing inside. Babel is the city humans build to consolidate their own power, erase difference, and put the system above the persons inside it. Jerusalem, in the document's reading, is the city ordered toward the common good, where persons remain ends and never become means. Both cities use technology. Both cities build. The difference is what the building is for, and who it is for.
For a CEO, this shifts the boardroom question. The question is no longer "should we adopt AI." That question has already been answered by everyone in your supply chain, every regulator with a mandate touching your industry, and every employee who is already using a chatbot in their workflow whether you authorized it or not. The question the encyclical raises cuts deeper: what kind of place are we becoming by how we adopt it? Are we building toward Babel or toward Jerusalem? That question reads as theological only in the narrowest sense. In any other read, it is the question any board with a fiduciary duty to a multi-stakeholder enterprise has to be able to answer.
The Babel-or-Jerusalem distinction is the encyclical's. The application of that distinction to enterprise AI deployment decisions is my read of it. Pope Leo XIV does not name a vendor selection criterion. He names a structural choice every institution making decisions about powerful technology has been making in some version since the eighteenth century. Reading that choice into the operating decisions a CEO is making this quarter is the work of applying the encyclical, not the work of summarizing it.
Why technology is never neutral
The encyclical's prior claim, also in paragraph nine, supplies the operational basis for the Babel-or-Jerusalem distinction.
Technology is never neutral, because it takes on the characteristics of those who devise, finance, regulate and use it.
That sentence is theologically grounded and operationally testable. Technology carries the values of its makers, its funders, its regulators, and its users into every deployment. The chatbot your customer service team is using carries the design decisions of the model provider, the safety choices of the policy team that fine-tuned it, the procurement choices of the vendor that resold it, and the prompt structures of the people who use it forty hours a week. Each of those layers transmitted values into the system you bought. Pretending otherwise is its own choice. The "we just use what works" posture is itself a values posture, and an unexamined one.
For executives deploying AI, that observation does the unglamorous but useful job of grounding values talk in operational language. Vendor selection is values selection. Procurement is values transmission. The acceptable-use policy you publish to your workforce is a values document whether or not anyone calls it that. The encyclical does not ask you to write a different mission statement. The encyclical asks you to recognize that the operating decisions you are making about AI right now are already values decisions, and to make them with that awareness.
The implication for the operator is direct. The vendor your company picks is going to shape how your employees experience AI for years. The training program your company funds is going to shape the proficiency floor of your workforce for the next decade. The deployment cadence you set is going to shape whether the workforce builds capability or builds resentment. None of these are neutral acts. They are values transmissions with downstream consequences. The encyclical's argument is that they should be treated that way explicitly, by the people in the room who hold authority over those decisions.
The dignity-of-worker question at the spine of the document
The encyclical's most direct claim to enterprise relevance sits in paragraph 37, which builds on Pope John Paul II's 1981 encyclical Laborem Exercens on human work.
the various kinds of job insecurity, fragmented career paths and automation must not be evaluated solely in terms of efficiency, but in relation to the dignity of the worker, the right to sufficient remuneration and the genuine possibility of participating in society.
That sentence cuts directly at the language most often used to justify AI investment cases. Efficiency gains, headcount reductions, productivity multipliers, throughput improvements: these are the standard inputs to an AI ROI conversation. The encyclical does not reject any of those measurements. It refuses the position that they are sufficient. Three co-equal criteria sit alongside efficiency in the document's weighting: dignity of the worker, sufficient remuneration, and participation in society. None is named as a soft factor or a stakeholder consideration sitting outside the core measurement.
Paragraph 51 names the failure mode directly.
Persons end up being reduced to a means of achieving results, a resource to be used and exploited, and are no longer recognized as a proper end.
In the same paragraph, the encyclical sharpens the principle behind the diagnosis.
The value of persons does not depend on what they achieve or produce. There are rights that apply to everyone simply by virtue of being human.
The failure-mode sentence is the strongest operator-side claim in the entire encyclical. It is the diagnosis of what goes wrong when efficiency-only evaluation becomes the only register an enterprise can speak in. People become inputs. The system gets optimized. The persons inside the system are measured against the optimization and judged on whether they keep pace. Worth gets attached to output. The encyclical names that as a structural failure of how the system was set up to evaluate itself, and locates the corrective at the level of how persons are valued rather than at the level of how productivity is measured.
The CEO who reads paragraph 51 honestly has to answer a hard question about their own organization. Do our measurement instruments for AI deployment have any non-efficiency variables in them? Do we have any way to surface whether a deployment treated the affected workers as ends or as means? In most companies I have worked with, the answer is no. Productivity dashboards are detailed. People dashboards are sentiment surveys taken twice a year. The asymmetry is the structural finding the encyclical is naming.
This is not a call to slow AI adoption. The encyclical is not anti-technology. The encyclical is asking a different question: when you adopt, what do you measure? The dignity question becomes operational the moment a measurement is added that surfaces it. The fix is not a values statement on the wall. The fix is a variable in the dashboard the leadership team actually looks at.
The technocratic critique
Paragraph 43 cites Francis, in Laudato Si' (2015), on the technocratic posture that reduces everything "to an object to be dominated." That citation pulls the encyclical's argument forward by adding the structural diagnosis of what goes wrong when an enterprise loses sight of the dignity question.
The encyclical is explicit that the question is not whether to use technology. The critique targets technocracy: a worldview in which efficiency, output, and instrumental control become the only valid criteria for evaluating systems and people. The distinction is load-bearing because it is the move most operators miss when they hear the encyclical's argument. Pope Leo XIV is not asking enterprises to use less AI. He is asking enterprises to stop evaluating themselves only with the language AI fluency lends them.
Inside an enterprise, the technocratic posture surfaces when AI procurement, AI deployment, and AI workforce planning all run through productivity-only criteria. The dignity question gets crowded out structurally, not because anyone willed it out. The criteria the leadership team uses to evaluate the work were built for an earlier era when measurement instruments for productivity were the hard part of the job. They were not designed to ask the dignity question. So the dignity question does not get asked. That is the structural finding the encyclical raises.
The fix is structural too. Add a non-instrumental variable to every AI evaluation. Rather than asking only "did this cut hours," ask "how did this affect the people who used to do this work." Rather than asking only "did this reach the throughput target," ask "what does the worker experience of this system look like at the six-month mark." The questions are not soft. They are measurable, and they belong in the same rubric that measures the hard numbers. The encyclical's argument is that until they are in the same rubric, they will keep losing to the variables that are.
The governance-of-data question (paragraph 72)
Paragraph 72 names a political-economy fact most enterprise AI conversations skip.
When it comes to decisions regarding economic flows and digital platforms, as well as the governance of data and algorithms, we cannot allow a handful of actors to dictate these processes on their own; instead, we must build forms of cooperation that respect the various levels of the global community and make them jointly responsible for the common good.
That sentence reads as policy commentary, and it is. It is also operationally relevant to every enterprise running its AI stack on infrastructure provided by a small number of vendors. A handful of model providers. A handful of cloud infrastructure providers. A handful of data brokers and training-data sources. A handful of policy frameworks shaping how those providers operate. The encyclical names the structural risk of that concentration, and it is correct to name it.
This is a call to refuse the assumption that the terms set by the largest actors are inevitable. The encyclical is realistic about the global political economy. The document does not pretend the concentration can be undone tomorrow. The document asserts that the concentration reflects choices rather than nature, and choices can be revisited.
For an enterprise, the practical reading is this. The vendor concentration in your AI stack is a strategic dependency, not a neutral fact of life. Pricing it in honestly is part of leading well. The diversification work that supply chain leaders do across other inputs is the same work AI architects should be doing across model providers, data sources, and inference layers. The reason is operational rather than moral. Concentration is a strategic exposure, and the encyclical is asking the right structural question about who holds it.
This connects to the AI governance work already happening at the policy level. The recent CARMA paper from RAND and the Future of Life Institute, which we covered at length in Performative AI Governance: 6 Tests From the New CARMA Paper, names the same structural risk from a different angle: institutional capture and decision-rights concentration as the primary failure modes for AI governance. The encyclical and the secular governance literature converge on the same observation. Concentration of decision rights inside a small number of actors is a structural risk that requires structural response. Two very different institutions, looking at the same operating environment, reaching the same diagnostic.
Where this maps in The 7 Levels of AI Proficiency
This section is the author's overlay. The encyclical does not name a proficiency framework. It does not endorse any specific assessment, training program, or organizational design. Reading the encyclical through The 7 Levels of AI Proficiency, the framework LaunchReady built, is my applied interpretation. It is one downstream attempt to operationalize the principles the document articulates. The encyclical is the canonical source. The 7 Levels of AI Proficiency is one operator's response to the dignity-of-worker question the encyclical raises.
With that boundary explicit, here is how the mapping holds.
Level 1, The Cadet (AI Aware), and Level 2, The Ensign (Prompt Engineer or Practitioner), are the levels where AI use is still skill-acquisition. The operator is learning to use the tool. The encyclical's dignity question exists at these levels but is not yet operational, because the person at this stage of proficiency is not yet making consequential decisions about how the system affects other people.
Level 3, The Lieutenant (Critical Thinker), is where the encyclical's question becomes operational. Critical thinking about AI use includes the criterion the encyclical names: am I treating the people affected by this deployment as ends or as means? A Level 3 operator is capable of asking that question and acting on the answer. Below Level 3, the question can be named but cannot reliably be answered, because the operator does not yet have the technical literacy to read the actual effects.
Level 4, The Commander (Context Engineer or Builder), is where the encyclical's principles start to shape system design. A builder choosing what to automate, what to keep human, what to flag for review, what data to expose, what data to protect, is doing Jerusalem-or-Babel work at the design layer. Every architectural choice transmits values. A Level 4 operator has the technical capability to make those choices deliberately rather than by default.
Level 5, The Captain (Design Thinker), and above is where the encyclical's principles become structurally embedded. A leader at this level is making procurement, hiring, training, and architecture choices that carry the values question into every layer. A Level 6 operator, The Admiral (Systems Integrator or Leader), and a Level 7 operator, The Mission Director (AI Orchestrator), are operating at the institutional level where the dignity question shows up in board-level discussions, in major capital allocation decisions, and in the design of the enterprise's relationship with its workforce.
The encyclical does not endorse this mapping. The mapping is mine. The mapping is offered as one possible operationalization of the values the encyclical articulates. Other operators will read the encyclical and operationalize it through different frameworks. That is appropriate. The encyclical is the canonical source; LaunchReady's work is one applied reading, and any leader applying the encyclical's principles should feel free to use whichever framework helps them.
How an Indianapolis CEO reads this on a Tuesday morning
Indiana sits inside the encyclical's argument in ways worth naming directly. The state has a strong Catholic intellectual tradition. The University of Notre Dame, in South Bend, is one of the country's leading Catholic research universities, and the Reilly Center for Science, Technology, and Values at Notre Dame does serious work at the intersection of science, technology, and the humanities. The Indiana Catholic Conference is the public-policy voice of the dioceses across the state.
For a Catholic CEO in Indianapolis, the encyclical is already on the way. It will arrive through the parish, through the diocesan paper, through the Indiana Catholic Conference's working communications. The text is going to enter their week through channels that are already part of their life. Their Tuesday morning question reads differently. The reading is already on the way; the open question is what to do with it.
For a Protestant CEO in Indianapolis, the encyclical arrives through the broader Christian tradition. Genesis is shared. The Babel reference sits inside the same scripture both traditions hold canonical. The dignity-of-work theology runs through both traditions, even when the ecclesial authority structures differ. The text is approachable; the principles are portable; the Protestant reader engages the document as one major Christian institution's serious thinking on a question every Christian executive is sitting with.
For a Jewish CEO, the encyclical's Babel-Jerusalem distinction is rooted in a scripture and a city tradition the Jewish reader knows from inside. The encyclical does not claim Jewish authority. The encyclical engages shared sources. A Jewish executive can read the encyclical as a serious cousin tradition naming the same question from its own location.
For a Hindu, Muslim, Buddhist, or secular CEO, the encyclical is a primary-source document from one of the most durable institutions in the world taking a serious position on the most consequential technology question of the decade. The institution does not ask the executive to convert. The institution asks the executive to take the question seriously. Read it the way you would read a McKinsey report from the only McKinsey that has existed for two thousand years. The credibility is not denominational. The credibility is institutional, historical, and intellectual.
Indiana CEO context completes the read. The state's AI policy environment is at the early stage of institutional build. The IN AI Initiative, announced in spring 2026, is the most visible structure. The state's narrow AI legislation passed and pending is the working policy surface, covered in detail in our Indiana AI Legislation 2026 Guide. Reading the encyclical alongside the state's working policy posture surfaces the same question both raise from different angles: who is the system for, and how do we know?
The Indianapolis CEO who reads the encyclical on a Tuesday alongside the ManpowerGroup AI confidence collapse data and the secular governance literature is reading three independent observers of the same operating environment, arriving at converging conclusions. That convergence is itself a signal.
Three things to do with the encyclical this week
The encyclical supplies a credible, historically grounded, independently authored standard for evaluating how an enterprise is deploying AI. Here is how to use it inside the company this week.
Read paragraphs 9, 37, 43, 51, and 72 directly.
The full text is at vatican.va. The five paragraphs named above carry the load. Twenty minutes of reading. The document is not technical; the language is plain. Reading it firsthand is faster than reading anyone else's summary, and the firsthand reading is what makes the document operational in your own thinking. Send the link to your leadership team and your board chair before the next AI conversation on the calendar.
Add a dignity variable to your next AI deployment review.
Pick the next AI deployment scheduled on your calendar. Add one non-efficiency criterion to the review: how does this deployment affect the dignity of the people who used to do this work, or who will be working alongside it now? The criterion does not have to drive the decision. It has to be visible in the record. Adding it to the review template is a one-meeting change. Keeping it in the template is the structural change that produces the operational effect the encyclical is asking for.
Audit your leadership team's proficiency, then audit how it routes the values question.
The 7 Levels of AI Proficiency assessment at assess.launchready.ai places each leader on a measurable scale in about ten minutes. The level is load-bearing because the encyclical's principles only become operational once the people running the decisions have enough proficiency to apply them to real choices. A Level 2 operator running a Level 5 question is a structural mismatch the encyclical's principles cannot fix on its own. Measure first, then design the team around the proficiency you actually have.
Related reading: What Is AI Proficiency: a Complete Guide and Performative AI Governance: 6 Tests From the New CARMA Paper and How to Start Using AI at Work: a 2026 Guide.
Sources
- Pope Leo XIV. Magnifica Humanitas: on Safeguarding the Human Person in the time of Artificial Intelligence. Holy See, May 15, 2026.
- Pope Leo XIII. Rerum Novarum. Holy See, May 15, 1891. Founding document of modern Catholic Social Doctrine on labor and capital.
- Pope John Paul II. Laborem Exercens. Holy See, September 14, 1981. Cited at paragraph 37 of Magnifica Humanitas.
- Pope Benedict XVI. Caritas in Veritate. Holy See, June 29, 2009. Cited at paragraphs 40 and 41.
- Pope Francis. Laudato Si'. Holy See, May 24, 2015. Source of the technocratic critique cited at paragraph 43.
- Pope Francis. Fratelli Tutti. Holy See, October 3, 2020. Cited at paragraph 44.
- Dignitas Infinita: Declaration on Human Dignity. Dicastery for the Doctrine of the Faith, April 8, 2024. Cited at paragraph 53.
- Reilly Center for Science, Technology, and Values. University of Notre Dame.
- Painter, Harrison. The 7 Levels of AI Proficiency. LaunchReady.ai.
- Performative AI Governance: 6 Tests From the New CARMA Paper. LaunchReady.ai Insights, May 23, 2026.
Frequently Asked Questions
What is Magnifica Humanitas and when was it issued?
Magnifica Humanitas is the first papal encyclical written specifically about artificial intelligence. Pope Leo XIV signed it on May 15, 2026, the 135th anniversary of Rerum Novarum, and the Vatican publicly presented it on May 25, 2026. The full title is on Safeguarding the Human Person in the time of Artificial Intelligence. The document runs more than seventy-five numbered paragraphs across the foundational opening chapters of its five-chapter structure, applying one hundred thirty-five years of Catholic Social Doctrine, starting with Leo XIII's Rerum Novarum in 1891, to the question of how the human person should be safeguarded as AI is deployed at scale across labor markets, public institutions, and daily life.
Who is the encyclical addressed to?
The encyclical is addressed to "all the Catholic faithful, to all Christians and to all men and women of goodwill". That phrasing is the established encyclical convention. The document is written to be read by anyone willing to engage the questions it raises about technology, work, and the human person. Catholic readers engage the document as binding Church teaching. Non-Catholic readers, including Christians from other traditions and readers from other religious or secular backgrounds, engage the document as a serious primary source from one of the most durable institutions in the world taking a position on the most consequential technology question of the decade.
What does the encyclical mean by "Babel or Jerusalem"?
In paragraph nine, Pope Leo XIV writes: "The primary choice is not between a yes or no to technology, but rather between constructing Babel or rebuilding Jerusalem." Babel and Jerusalem are biblical cities the encyclical uses to mark the two ends of a moral geography. Babel is the city humans build to consolidate their own power, erase difference, and put the system above the persons inside it. Jerusalem, in the document's reading, is the city ordered toward the common good, where persons remain ends and never become means. The encyclical's argument is that both cities use technology; the difference is what the building is for, and who it is for. Applied to enterprise AI, the distinction reorganizes the boardroom question from "should we adopt AI" to "what kind of place are we becoming by how we adopt it."
Why does the encyclical say technology is "never neutral"?
In paragraph nine, the encyclical states: "Technology is never neutral, because it takes on the characteristics of those who devise, finance, regulate and use it." The claim is theologically grounded but operationally testable. Every layer of a technology stack, from the people who design it to the funders who pay for it to the regulators who shape it to the users who deploy it, transmits values into the system. The "we just use what works" posture is itself a values posture. Applied to enterprise AI, this means vendor selection is values selection, procurement is values transmission, and the acceptable-use policy your workforce reads is a values document whether or not anyone calls it that.
What does the encyclical say about job insecurity and automation?
In paragraph 37, building on Pope John Paul II's Laborem Exercens (1981), the encyclical states: "the various kinds of job insecurity, fragmented career paths and automation must not be evaluated solely in terms of efficiency, but in relation to the dignity of the worker, the right to sufficient remuneration and the genuine possibility of participating in society." Paragraph 51 names the failure mode directly: "Persons end up being reduced to a means of achieving results, a resource to be used and exploited, and are no longer recognized as a proper end." In the same paragraph, the encyclical sharpens the principle behind the diagnosis: "The value of persons does not depend on what they achieve or produce. There are rights that apply to everyone simply by virtue of being human." The encyclical does not reject efficiency measurement. It refuses the position that efficiency measurement is sufficient. Three criteria are named co-equal with efficiency: dignity, sufficient remuneration, and participation in society.
What is the "technocratic" critique the encyclical makes?
In paragraph 43, citing Francis (Laudato Si', 2015), the encyclical critiques the technocratic posture that reduces everything "to an object to be dominated." The critique targets technocracy as a worldview rather than technology as a tool. The encyclical is explicit that the question is not whether to use technology. The critique targets the disposition in which efficiency, output, and instrumental control become the only valid criteria for evaluating systems and people. Inside an enterprise, technocracy surfaces when AI procurement, AI deployment, and AI workforce planning all run through productivity-only criteria; the dignity question gets crowded out structurally because the criteria the leadership team uses were not designed to ask it.
How does this apply to a CEO who is not Catholic?
The encyclical is addressed to "all the Catholic faithful, to all Christians and to all men and women of goodwill," which is the established encyclical convention for documents written for both Catholic and non-Catholic audiences. A non-Catholic CEO, whether Protestant, Jewish, Muslim, Hindu, Buddhist, or secular, can engage the document as a serious primary source from one of the most durable institutions in the world. The institution does not ask the executive to convert. The institution asks the executive to take the question seriously. The Babel-Jerusalem distinction is rooted in scripture shared across the Abrahamic traditions; the dignity-of-work theology has parallels across most major religious traditions; the political-economy analysis applies regardless of religious commitment. Read the encyclical the way you would read a credible primary source from any institution with two thousand years of operating history thinking carefully about the question every executive is now sitting with.
How does this connect to The 7 Levels of AI Proficiency?
The encyclical does not name a proficiency framework and does not endorse any specific assessment, training program, or organizational design. The connection is an applied overlay by the author: one downstream attempt to operationalize the principles the encyclical articulates. The mapping holds that Level 1 (The Cadet) and Level 2 (The Ensign) are skill-acquisition levels at which the dignity question exists but is not yet operational; Level 3 (The Lieutenant, Critical Thinker) is where the question becomes operational in daily decisions; Level 4 (The Commander, Context Engineer or Builder) is where the principles start to shape system design; Level 5 (The Captain, Design Thinker) and above is where the principles become structurally embedded in procurement, hiring, training, and architecture decisions. The mapping is the author's read of the encyclical applied to LaunchReady's framework; other operators will read the encyclical and operationalize it through different frameworks. The encyclical is the canonical source; LaunchReady's work is one applied reading.
