There is a pattern worth observing, one that repeats across centuries with remarkable consistency. The Catholic Church, as an institution, has repeatedly shown a strong ability to position itself alongside whoever holds power at a given moment. This is not a moral judgment but a historical observation, and it becomes especially relevant when we look at the Vatican’s recent foray into artificial intelligence discourse.

cover

Before we discuss algorithms and neural networks, let us take a brief tour through a few historical moments. These are not episodes from distant centuries, but modern examples that help clarify the pattern.

A consistent pattern of proximity to power

Consider Pope Benedict XV, who during World War I maintained a studied neutrality that allowed the Vatican to maintain relationships with both the Central Powers and the Allied forces. While this neutrality was presented as a moral stance above the fray, it had the practical effect of ensuring the Church maintained influence regardless of which side emerged victorious. The pattern was clear: position yourself so that whoever wins, you remain relevant.

The relationship between Pius XI and Mussolini provides an even more striking example. The 1929 Lateran Pacts resolved the “Roman Question” that had plagued Church-State relations since Italian unification, but at what cost? The Vatican gained sovereign territory, financial compensation, and recognition of Catholicism as Italy’s state religion. In exchange, it provided legitimacy to a fascist regime. The Church got what it needed from power, and power got what it needed from the Church.

Then there is the troubling case of Pius XII and the Nazi regime. The extent of Vatican involvement in facilitating escape routes for Nazi war criminals after World War II remains debated, but the documented evidence is enough to show that institutional survival and influence often took precedence over justice. Studies such as Gerald Steinacher’s Nazis on the Run reconstruct how clerical and diplomatic networks intersected with postwar escape routes. The so-called “ratlines” that helped Nazi officers escape to South America were not purely humanitarian operations; they were also strategic calculations about maintaining influence in a rapidly changing geopolitical landscape.

More recently, we saw John Paul II’s relationship with Chilean dictator Augusto Pinochet. While the Pope spoke eloquently about human rights in Poland and Eastern Europe, his stance toward Latin American dictatorships was considerably more accommodating. The famous 1987 photograph of John Paul II appearing on a balcony with Pinochet during a period of brutal repression speaks volumes about institutional priorities.

The pattern is stable across different eras. When there is power to be negotiated with, the Church positions itself as an essential interlocutor. When there are moral compromises to be made for institutional advantage, those compromises are made with remarkable pragmatism, often wrapped in theological language that obscures the transactional nature of the relationship.

Enter the encyclical on artificial intelligence

Which brings us to the present moment and the Vatican’s recent intervention on artificial intelligence, including the document Antiqua et nova. On the surface, this appears to be a theological and ethical reflection on emerging technology. The text discusses human dignity, the risks of algorithmic bias, the importance of human oversight, and the need for ethical frameworks in AI development. All sensible positions that few would dispute.

But if we understand the historical pattern, we should ask a different question. This is not primarily a theological statement. It is a strategic positioning move that presents the Church as a necessary ethical arbiter in a conversation where democratic institutions have largely failed to establish coherent governance frameworks.

Think about the current state of AI regulation. Governments are struggling. The European Union has produced the AI Act, which is comprehensive, extraordinarily complex, and already facing implementation challenges. The United States has offered little coherent federal framework, leaving the field to a patchwork of state regulations and voluntary industry commitments. International coordination is minimal. The technology is evolving faster than policy can keep pace.

Into this vacuum steps the Vatican, offering moral clarity and ethical frameworks. It is a brilliant move, not because the encyclical is technically groundbreaking, but because it establishes the Church as a legitimate voice in a conversation where legitimacy is in short supply.

As I explored in my analysis of how organizations can adopt AI security tools without losing control, the governance gap in AI deployment is real and consequential. The Church is simply doing what it has always done: identifying where power is accumulating and positioning itself as an essential participant in that power’s exercise.

Understanding the Overton Window

To fully appreciate what is happening here, we need to understand a concept from political theory called the Overton Window. Named after policy analyst Joseph P. Overton, the Overton Window describes the range of policies and ideas that are politically acceptable to the mainstream population at a given time. Ideas outside this window are considered too radical, too extreme, or simply unthinkable in current political discourse.

The window is not fixed. It can be shifted, expanded, or contracted through deliberate effort. Think of it as a frame that determines what kind of conversation is even possible in public discourse. Ideas inside the window can be debated, refined, and potentially implemented. Ideas outside the window are dismissed without serious consideration, regardless of their actual merit.

For a concrete example, consider the debate around end-to-end encryption and government access. For years, the Overton Window on this issue included everything from “strong encryption is essential for security and privacy” to “backdoors for law enforcement are necessary.” What shifted the window was not technical arguments but strategic framing that redefined what counted as a “reasonable” position.

When a powerful institution issues a formal position on an emerging issue, it is not merely stating an opinion. It is attempting to define the boundaries of the Overton Window for that issue. In practical terms: “Here is the range of acceptable positions. Arguments outside this range are extreme, unreasonable, or morally problematic.”

The Vatican’s AI encyclical does exactly this. It establishes parameters: AI should respect human dignity, should include human oversight, should be deployed with ethical considerations, should not replace human judgment in critical decisions. These positions are framed as moral imperatives rooted in centuries of theological tradition.

Notice what this framing accomplishes. It positions certain arguments as inherently more legitimate (those consistent with Church teaching) and others as problematic (those that treat AI as fully autonomous or that minimize the importance of human dignity). It does not ban certain technical approaches, but it does make them morally suspect, which in many contexts is more effective than an outright prohibition.

The Church as necessary interlocutor

The encyclical makes an implicit but crucial argument: democratic states have failed to regulate AI adequately, therefore moral and religious institutions must fill the gap. This is presented as a service to humanity, a necessary intervention to prevent technological development from proceeding without ethical guardrails.

But let us be clear about what this claim actually means. It asks us to treat an institution with no democratic accountability, no technical expertise in AI systems, and no formal authority over technology policy as a legitimate, even necessary, participant in shaping how AI is governed globally.

Why should we accept this framing? The answer provided by the encyclical is essentially: because we have been thinking about human dignity and ethical questions for two thousand years. This is not an unreasonable claim on its face. The Church does have a long tradition of moral philosophy and ethical reasoning.

But history suggests we should be skeptical. When the Church positions itself as an ethical arbiter, we should ask: whose interests are actually being served? The answer is rarely as simple as “humanity’s interests” or “ethical principles.” More often, the answer is: the institutional interests of the Church itself, which involve maintaining relevance and influence in a rapidly secularizing world.

At the same time, the strongest version of the opposing argument deserves to be stated clearly. In a fragmented policy landscape where tech firms move faster than legislatures, institutions with long ethical traditions can provide a shared vocabulary that travels across borders. Even critics of clerical influence can acknowledge that principles like dignity, proportionality, and accountability may help slow the drift toward purely market-driven AI deployment.

Consider the parallel with Rerum Novarum, Pope Leo XIII’s 1891 encyclical on the condition of workers. That document is often celebrated as an early statement of social justice principles, addressing the exploitation of workers during the Industrial Revolution. And indeed, it did contain genuine concern for working conditions and labor rights.

But it also contained something else: a vigorous argument against socialism and a defense of private property as a natural right. The encyclical was not merely about helping workers. It was about ensuring that the response to industrial capitalism did not take the form of socialist revolution, which would have threatened both economic elites and the Church’s own institutional position. It offered a “third way” that preserved existing power structures while making them marginally more palatable.

The AI encyclical functions similarly. It acknowledges real concerns about algorithmic bias, automation’s impact on employment, and the concentration of technological power. But its proposed solution is not democratic governance, worker control of technology, or the breaking up of tech monopolies. Instead, it offers ethical frameworks, human dignity principles, and the suggestion that institutions like the Church should be consulted in AI governance.

The trap for democratic institutions

Here is where the situation becomes genuinely interesting from a governance perspective. Politicians and policymakers who are struggling with AI regulation now have an attractive option: adopt the Church’s ethical framework as a foundation for policy. This offers several advantages.

First, it provides moral legitimacy. A policy based on principles endorsed by major religious institutions can be presented as something more than technocratic rulemaking. It becomes a defense of human values against the excesses of technology.

Second, it offers political cover. If a policy fails or proves inadequate, the blame can be shared with the religious institutions that provided the ethical framework. “We followed the moral guidance provided by respected institutions” is a useful defense against criticism.

Third, it avoids the hard work of understanding the technology and its implications. Rather than developing technical expertise and crafting nuanced regulations based on how AI systems actually work, policymakers can outsource ethical framing to institutions that claim moral authority.

The trap is that this approach substitutes moral declaration for technical governance. As I discussed in my piece on human washing and AI ethics, declaring a commitment to human dignity and ethical AI is not the same as implementing governance structures that ensure AI systems are developed and deployed responsibly. It risks becoming what we might call “ethics washing,” where moral language obscures the absence of meaningful regulation.

The question is whether policymakers will recognize this trap or whether they will gratefully accept the Church’s offer to help define the terms of AI governance. The answer will likely vary by country and political context. In heavily Catholic countries, the temptation to incorporate Church teaching into AI policy will be strong. In more secular contexts, the language might be softened but the underlying framework could still be influential.

What history teaches us to watch for

If we take the historical pattern seriously, what should we be watching for in the coming months and years? Several things.

First, watch for the Church seeking formal advisory roles in national and international AI governance bodies. This might take the form of consultative status in UN organizations, advisory positions in national ethics committees, or formal partnerships with governments developing AI strategies.

Second, watch for the encyclical’s language appearing in policy documents and legislation. When politicians start talking about “human dignity” and “ethical frameworks” in the context of AI, ask whether they are simply using the Church’s vocabulary without acknowledging the source. This matters because it represents a subtle shift in who is defining the terms of the debate.

Third, watch for the Church’s positions on AI becoming a litmus test for respectability in technology ethics discussions. If anyone proposing AI governance approaches that differ significantly from the encyclical’s framework is labeled as ethically irresponsible or dismissive of human dignity, that is the Overton Window in action.

Fourth, and perhaps most importantly, watch for what is not discussed. The encyclical focuses heavily on individual dignity and human oversight, but it says relatively little about questions of power, ownership, and democratic control of AI systems. This is not an oversight. It reflects institutional priorities that align more closely with existing power structures than with radical democratization of technology.

The pattern from Pius XI to John Paul II suggests that when the Church positions itself with power, it does so in ways that preserve rather than challenge fundamental hierarchies. The AI encyclical likely represents a similar calculation: stake a claim in AI governance that maintains the Church’s relevance without threatening the interests of the tech companies and governments that will ultimately control these systems.

For policymakers, the practical question is not whether to accept or reject religious ethical input wholesale. The practical question is how to translate values into enforceable rules. A credible AI governance agenda should include independent auditing powers, clear liability regimes for high-risk deployments, mandatory transparency on training data provenance, and conflict-of-interest safeguards for advisory bodies. Without these mechanisms, ethical language remains performative, no matter who articulates it.

A gentle skepticism

None of this means the encyclical’s concerns about AI are illegitimate or that all of its recommendations are problematic. Many of the ethical principles it articulates, respect for human dignity, the importance of transparency, the need for accountability, are genuinely important and worth defending.

But we should understand what we are looking at. This is not primarily a theological document. It is a strategic positioning move by an institution that has spent centuries perfecting the art of aligning itself with power while claiming to speak for higher principles.

The irony is that the Church’s historical record on embracing new technologies is, in several respects, more progressive than its reputation suggests. What it has been consistently excellent at, however, is preserving a role in defining the ethical boundaries around those advancements.

Will policymakers substitute the Church’s moral framework for technical governance? Or will they develop the expertise and political will to craft AI regulations based on how these systems actually work and what forms of democratic oversight are genuinely needed?

History suggests pessimism is warranted, but history also contains surprises. Perhaps this time will be different. Perhaps policymakers will recognize that AI governance requires technical expertise, democratic accountability, and genuine engagement with questions of power and control, not just ethical frameworks provided by institutions with no democratic mandate.

Or perhaps the pattern will continue, and a few years from now we will find ourselves discussing how the Church managed to position itself as an essential voice in AI governance without ever having to demonstrate technical expertise or submit to democratic accountability. In which case, we will have witnessed not a moral intervention but a familiar performance: institutional power recognizing where new forms of power are emerging and ensuring it maintains a seat at the table.

The Church has been playing this game for centuries. It is very, very good at it. Recognizing the pattern is the first step toward building AI governance that is democratic, technically grounded, and accountable.