Pope Leo Drops 42,000-Word AI Encyclical With Anthropic Co-Founder at His Side

On Monday, Pope Leo XIV made history — not just by releasing the Catholic Church first‑ever encyclical on artificial intelligence, but by doing so alongside a Silicon Valley billionaire. Christopher Olah, co‑founder of ⁵ and the company interpretability research lead, sat just a few seats from the pontiff in the Vatican Synod Hall as Leo read aloud from Magnifica Humanitas — Latin for magnificent humanity.

The 42,300‑word document is the heaviest form of papal teaching there is, and TIME reports it marks Leo most sweeping statement yet on the promise and dangers of AI. The pontiff formally signed the encyclical on May 15 — exactly 135 years after Pope Leo XIII Rerum Novarum, the landmark text that addressed the Industrial Revolution impact on workers.

The Vatican doesn normally invite outsiders to speak, let alone those in the tech industry, Vanity Fair noted. A senior Vatican source told the National Catholic Reporter that Olah invitation signals the great willingness for us to participate more fully in the dialogues that are going on.

The encyclical is not a gentle suggestion. The Guardian reports that Leo denounced a culture of power driving AI development, warning that governments must not leave the technology future to be shaped solely by private companies or the invisible hand of the market.

The document hits several specific targets:

The Vatican did not pick Anthropic at random. ⁵ traced the relationship back to Anthropic founding, when Dario and Daniela Amodei left OpenAI in 2021 convinced that AI models were becoming too powerful to be developed exclusively according to the logic of competition and speed.

Anthropic public positioning around Constitutional AI — training models against an explicit set of ethical principles rather than just correcting dangerous outputs — aligned with the Vatican interest in AI alignment. The Vatican began engaging seriously with AI in 2020 through the Rome Call for AI Ethics, an initiative with Microsoft, IBM, and others. But as ChatGPT and the US‑China AI race accelerated, the Holy See concluded the issue was no longer just about tech ethics, but about the very future of humanity, WIRED reported.

Olah specifically was chosen because he represents the more theoretical and almost philosophical side of AI research, WIRED noted. While CEO Dario Amodei handles media and policy, Olah leads interpretability — the effort to understand what actually happens inside neural networks. Vanity Fair reports Olah was raised evangelical Christian before becoming an atheist at 15, and that his personal writings emphasize moral connection and his frustratingly scrupulous dedication to doing what he believes is right.

The encyclical landed in the middle of an already‑heated Washington debate over AI regulation. CNBC reports that Interior Secretary Doug Burgum dismissed Leo warning outright, telling Fox Business: I didn know that tech editorializing was part of the role of being pope.

But Vice President JD Vance — the highest‑ranking Catholic in the administration — called the same message profound and praised it as the kind of moral leadership the church should offer at the start of the AI age, according to CNBC.

The split comes just one week after Trump delayed an executive order that would have created a voluntary AI safety review process, citing concerns about maintaining the US edge against China. The timing created a dilemma for a president who won 55% of Catholic voters in 2024, CNBC reported. Political scientist Ryan Burge of Washington University told CNBC that repeated clashes with Leo over immigration, war, and now AI could erode support among less firmly committed Catholic voters in swing districts ahead of the 2026 midterms.

Olah speech, published in full on Anthropic blog, went further. There is a real possibility that AI will displace human labour at very large scale, he said. If that happens, supporting those displaced will be a moral imperative of historic proportions. This was not the standard tech‑industry line about AI creating more jobs than it destroys — it was a frank acknowledgment, delivered in the Vatican, that the damage could be real and massive.

Olah added that companies like his operate inside a set of incentives and constraints — commercial, geopolitical, and personal pressures — that can conflict with doing the right thing, The Guardian reported. We need informed critics who will tell the labs when we are failing, Olah said. We need moral voices that the incentives cannot bend.

The encyclical by itself changes no laws. But it changes the conversation — and it does so at a moment when the regulatory window is both open and closing simultaneously.

For developers and AI builders, three signals stand out. First, the Vatican-Anthropic alliance creates a new axis of AI governance pressure that does not depend on Washington. Europe AI Act already exists. Now the Catholic Church — an institution with 1.4 billion members and diplomatic relationships with nearly every country — has entered the arena with its most authoritative teaching document. Second, Olah concession that even safety‑focused companies cannot self‑regulate undercuts the tech industry primary argument against regulation. When an Anthropic co‑founder tells the Pope that outside scrutiny is essential, the trust us line becomes harder for any company to sell. Third, the encyclical specific focus on worker displacement and autonomous weapons gives labor groups and arms‑control advocates new moral ammunition. CNBC notes that Democrats and AI‑safety advocates could use Leo warning to argue the administration is too deferential to Silicon Valley and too dismissive of concerns about workers, families and national security.

No one is rewriting their API terms of service based on a papal document. But the encyclical shifts the Overton window — making it harder for governments to claim AI regulation is unnecessary, and harder for companies to claim it impossible.