Anthropic was expelled by the United States and then found refuge in Rome—this story sounds like something out of the Middle Ages.

Author and source: 0x9999in1, ME News



TL;DR

• Pope Leo XIV and Anthropic co-founder Christopher Olah jointly unveiled the first AI encyclical, announcing a formal partnership between the Vatican and Anthropic focused on AI ethics and the future of humanity.
• This is the most symbolic international political breakthrough following the full ban on Anthropic by the Trump administration.
• Vatican's choice of Anthropic over OpenAI or Google sends a clear signal of alignment with a "safety-first" approach.
• Anthropic's revenue in Europe has increased nearly tenfold over the past year, with its political strategy deeply intertwined with its business expansion.
• Peter Thiel’s accelerationist faction met with cold reception in Rome, as the Vatican’s stance on technology ethics becomes increasingly clear.
• This partnership is essentially a two-way "moral arbitrage"—Anthropic needs legitimacy, and the Vatican needs technological influence.
• Dario Amodei is set to visit Rome and London, revealing plans for a data center deployment in Southern Europe.

A banished company entered the world's oldest center of power.

May 2026.

St. Peter's Square in the Vatican. Pope Leo XIV, dressed in a white liturgical robe, stands beside a young man in a plaid shirt—Christopher Olah, co-founder of Anthropic and a pioneer in neural network interpretability.

The screen itself is a statement.

An AI company designated by the U.S. government as a "security risk" had its lead scientist stand at the heart of the Catholic world, alongside the spiritual leader of 1.3 billion believers, to release the first papal encyclical in human history on artificial intelligence.

This is not a charity event. Not an academic seminar. Not a typical PR stunt by a tech company.

This is a politically precise alignment to the millimeter.

Let’s go back three months: At that time, Anthropic was a company under full-scale crackdown in its home market. In February, the Trump administration signed an executive order requiring all federal agencies to cease using Anthropic’s technology. The Department of Defense unprecedentedly added a domestic tech company to its list of security risks. Why? Anthropic refused to remove the safety red lines in its AI system regarding lethal autonomous weapons.

In other words—because this company said "no".

We do not build killing machines. We do not assist the military in bypassing ethical reviews. We do not create unguarded general-purpose weapons.

What was the cost? Within hours of the ban order, the Pentagon contract was taken over by OpenAI.

So when Ola stands beside the Pope, the subtext of this image is crystal clear:

You don't want me? That's fine. Someone else does. And this "someone" leads the world's largest moral authority.

Why the Vatican? Why now?

Answer the first question first.

The Vatican is not a cutting-edge technological institution. It has no data centers, no GPU clusters, and may not even have a proper IT department. But it has one thing that no tech company can buy with money—moral legitimacy.

In the global competition over AI governance, the question of "who decides" remains unresolved. The U.S. government wants to call the shots, but Trump’s "remove the guardrails" policy has raised alarms worldwide. The EU wants to lead, but its AI Act moves slowly and lacks technical capacity. China wants to assert control, but deep geopolitical trust deficits stand in the way.

What about the Vatican? It doesn’t manufacture chips, train models, or participate in commercial competition. Yet it holds undisputed spiritual authority over 1.3 billion Catholics worldwide and wields significant soft power in the Global South—Africa, Latin America, and Southeast Asia.

More importantly: the position of Leo XIV himself.

The new pope who assumed office in 2025 has repeatedly spoken out publicly about AI risks since taking office. He is not a political figure who vaguely advocates for "technology for good"—he has named and criticized "unconstrained technological accelerationism" and cited specific AI safety research papers. His predecessor, Francis, began addressing AI ethics in his later years, while Leo XIV has elevated it to a top priority in Vatican diplomacy.

Now, consider the timing.

What will the global AI landscape look like in the first half of 2026? In the United States, the safety camp is being suppressed, while the acceleration camp dominates. OpenAI secures defense contracts, deepening its collaboration with the Department of Defense to an unprecedented level. Google DeepMind’s safety team continues to lose talent. Meta has outright announced the abandonment of all external safety audits.

The prevailing sentiment across Silicon Valley is that "speed" overrides "stability".

In this context, the Vatican’s decision to establish a formal partnership with Anthropic—the only company that has genuinely paid a financial price for upholding safety boundaries—is not a casual one. It is a deliberate and thoughtful alignment.

Tier's cold reception and Ola's red carpet

There is a detail worth reflecting on repeatedly.

Just a few weeks ago, Peter Thiel—co-founder of PayPal, CEO of Palantir, and a key tech advisor to Trump—flew to Rome personally to pitch the Silicon Valley accelerationist vision to the Vatican.

What’s the result?

Cold response. No public events, no joint statements, no papal audience. Thiel flew to the Vatican with a stack of PPTs on "technology liberating human potential," then quietly returned to the United States.

Compare the treatment Olah received: co-releasing an encyclical, global live streaming, a formal cooperation framework, and personal endorsement by the Pope.

Placing these two images together sends a clearer signal than any diplomatic language—

The Vatican is not just choosing partners. It is choosing sides.

Which side to choose? Choose the side that says "AI should have boundaries." Choose the side that says "Not all technological progress is worth pursuing." Choose the side that says "Human dignity cannot be overwritten by algorithms."

You could call it conservatism. You could call it technophobia. But when the world’s most morally influential institution makes this choice, its ripple effects far surpass any government white paper.

Arbitrage: What did Anthropic receive, and what did the Vatican receive?

Let’s be honest.

This is not a purely idealistic alliance. Both sides have their own calculations, and both are playing them shrewdly.

First, moral legitimacy.

Being banned by its home government means that for a company, every business negotiation in the global market will be met with the same question: "Do you have any issues?"

The U.S. government labeling you as a security risk—even if the reasoning is absurd—is poison on the international market.

But what if the Pope said: I trust this company, I’m partnering with them?

This is no longer just endorsement. It’s the highest form of moral whitewashing—no, it’s moral legitimization. Because Anthropic was banned precisely for upholding ethical standards, the Vatican’s endorsement is logically consistent.

Second, the political gateway to the European market.

Anthropic's revenue in Europe has increased nearly tenfold over the past year. This figure is impressive, but sustained growth requires more than just product strength. Europe is the world’s most strictly regulated AI market, and the AI Act has just entered its enforcement phase, with countries actively vetting "trustworthy AI providers."

With the Vatican’s partnership and endorsement, Anthropic is no longer just an American tech company in Catholic-majority Southern European markets—Italy, Spain, and France—but has become a "Vatican-certified" benchmark for AI ethics.

This provides tangible support for securing government contracts, obtaining data center approvals, and gaining public acceptance.

Third, the gateway to the Global South.

Africa, Latin America, the Philippines—policymakers in these regions are naturally wary of Silicon Valley giants but have an inherent affinity for the Vatican. If Anthropic wants to enter these markets, its relationship with the Vatican is the best possible introduction.

First, technological authority.

The Vatican can issue a hundred documents on AI ethics, but without establishing substantive partnerships with any leading AI companies, these documents are merely theoretical.

Partnering with Anthropic means the Vatican’s ethical framework has the opportunity to be embedded into real AI system designs. It means when the Pope says, "AI should respect human dignity," a company valued at hundreds of billions of dollars is responding to that call with actual technical architecture.

This is a leap from chanting slogans to taking action.

Second, geopolitical independence.

A core challenge facing Leo XIV is maintaining the Vatican’s neutrality and independence amid the U.S.-China tech rivalry. Fully aligning with the U.S. mainstream tech camp (OpenAI, Google, Meta) would risk portraying the Vatican as an appendage of U.S. technological hegemony.

But Anthropic is different. It has already been severed by the U.S. government. It is a company that is "American, but does not represent the position of the U.S. government." By partnering with it, the Vatican can access cutting-edge technology without bearing the cost of political affiliation.

Third, the attention of the younger generation.

Frankly, the global Catholic Church is facing a serious decline in young adherents. AI ethics happens to be an issue of high concern among young people—especially those with higher education. The Vatican’s proactive engagement on this issue is a forward-looking strategy for retaining congregants.

The weight of an encyclical

Let’s talk about the encyclical itself.

An encyclical holds a prominent position within the Catholic Church as an official teaching document issued by the Pope with his highest authority, second only to dogmatic constitutions. Historically, significant encyclicals are few in number—the 1891 Rerum Novarum initiated the Catholic social teaching tradition, and the 1967 Populorum Progressio shaped the framework for global development aid.

Now, AI is receiving the same treatment.

This means that, within the theological and ethical framework of Catholicism, AI issues have been elevated to the same level of significance as labor rights, wealth inequality, and war and peace. Over two hundred thousand Catholic churches, five thousand Catholic universities, and countless affiliated hospitals and charitable organizations will be called upon to take seriously the guidance in this document.

Scientists from Anthropic were involved in drafting this document.

What does this mean? It means a paradigm of technological-ethical integration is emerging: not technology companies building products first, followed by ethicists playing catch-up with criticism; but rather ethical frameworks and technical design advancing in tandem, inherently linked from the outset.

You can question the practical effectiveness of this model. But at least symbolically, it breaks the industry’s ingrained habit of “technology first, ethics afterward.”

European offensive: From morality to business

Morality is one thing, business is another.

Anthropic has appointed Liam Booth-Smith, former chief of staff to the UK Prime Minister, to lead its European expansion—this hire alone speaks volumes: the company needs not just a sales director, but a political operator who can navigate between Downing Street and the Élysée Palace.

Dario Amodei's itinerary has been revealed: after Rome, he will fly to London to meet UK Prime Minister Starmer. Topics include a cooperation framework with the UK AI Safety Institute, data center location, and potential investment in a sovereign AI fund.

The data center project in Southern Europe is more worth noting. In recent years, the Italian government has actively promoted digital infrastructure development, and if Anthropic can establish a data center in Italy, it would not only mean a compliance advantage with EU data sovereignty regulations but also signify establishing a forward base across the Mediterranean region.

This timeline precisely coincides with the announcement of the collaboration with the Vatican.

Coincidence? Of course not.

This is a comprehensive strategy: first secure moral endorsement, then discuss political cooperation, and finally implement commercial projects. Each step paves the way for the next, with a logical chain as clear as a textbook case.

The last bastion of the security camp or a new beginning?

Take a step back to see the bigger picture.

In 2026, the AI industry is in a strange state of division.

On one side is the acceleration camp centered on the United States: regulatory deregulation, full opening of military applications, reduced funding for security research, and the catchphrase "If we don’t do it, China will" serving as a universal excuse.

On the other side is a loose but increasingly structured global safety alliance: the EU’s AI Act, the UK’s AI Safety Institute, Canada’s AISI, Japan’s AI Safety Network… and now the Vatican.

Anthropic is trying to position itself as the technological core of this alliance.

This is a bold bet. It bets that, in the long run, safety is not a barrier to AI development, but a prerequisite for AI’s acceptance by society. When public fears and mistrust of AI eventually translate into political pressure—this bet assumes that day will come—companies that took safety seriously from the start will gain the greatest market access advantage.

Is this bet correct?

To be honest, no one can give a definitive answer right now. If the accelerationists are right, if AGI truly arrives within the next two to three years, and if "speed" is truly the only strategy… then Anthropic’s choice may indeed be a form of self-marginalization.

But what if the "slow" camp is right? What if social integration of AI takes ten years or longer? What if public backlash truly arrives?

The seeds Anthropic planted today in the Vatican could become the greatest strategic asset in the future.

An uncomfortable question

Finally, we must raise a somewhat uncomfortable question.

Is there risk in a commercial company being deeply tied to a religious authority?

Yes.

The risk is that once moral authority is infiltrated by commercial interests, its credibility is undermined. If, in the future, Anthropic makes a controversial decision on a specific issue—such as serving an authoritarian government or compromising on safety standards—the Vatican’s endorsement could become a liability.

Conversely, if the Vatican’s ethical guidance becomes too detached from technological reality and turns into abstract moralizing, Anthropic may gradually reduce this relationship to a mere public relations tool.

This tension is real and the most important variable to monitor over the long term in this partnership.

But at least today, at this specific historical moment—a company exiled for upholding its principles and an institution built on moral authority have fulfilled a precise mutual need—this alone carries more weight than most AI governance initiatives.

Conclusion

Return to the scene at St. Peter's Square.

Ola wears a plaid shirt, the Pope wears a white robe. A scientist studying the interpretability of neural networks, an elderly man claiming to represent God.

Absurd? Maybe.

But in an era where all old orders are crumbling—where governments are failing, international institutions are stagnant, and corporate self-regulation has become a joke—perhaps it is precisely this seemingly absurd combination that can pry open even a small amount of real change.

Anthropic was expelled by the United States and found refuge in Rome.

This story sounds like something from the Middle Ages.

But this is 2026.

Reference:

1. Executive Order on Federal AI Procurement Standards, The White House, February 2026.
2. "Pentagon Designates Anthropic as Security Concern, Awards Contract to OpenAI," The Washington Post, February 2026.
3. Pope Leo XIV, Encyclical on Artificial Intelligence and Human Dignity, Vatican Press, May 2026.
4. "Anthropic Hires Former UK PM Chief of Staff for European Expansion," Financial Times, March 2026.
5. "Peter Thiel's Rome Visit Fails to Win Vatican Backing for AI Acceleration,"Politico Europe, April 2026.
6. European Commission, "AI Act Implementation: First Compliance Reports," Brussels, Q1 2026.
7. "Anthropic European Revenue Growth Surges Tenfold," Bloomberg , April 2026.
8. "Dario Amodei Plans Rome and London Diplomatic Tour,"Reuters, May 2026.