The Pope, Anthropic, and the battle for "trust" in the AI industry.

There's a 33-year-old Canadian who recently got the most expensive PR moment in the history of his industry. On top of that, he got it for free, and his name is Christopher Olah, and he was standing recently on stage at the Vatican next to the Pope. For context, Olah is the co-founder of Anthropic, one of the most valuable AI companies on the planet, with some value around $900 billion, and, uh, a couple of days ago, he he walked into the Synod Hall in the Vatican to release, together with Pope Leo XIV, the first papal encyclical in history on artificial intelligence: Magnifica Humanitas, which in Latin stands for Magnificent Humanity. It was actually released exactly 135 years after the other Pope, Leo XIII, wrote his encyclical on the Industrial Revolution and workers' rights. That was called Rerum Novarum. But on May 25th, 2026, there have been two firsts. For the first time in history, a pope personally presented an encyclical to the world. Usually this gets delegated to cardinals. Leo XIV went to the Synod Hall himself, and among the invited speakers was also Christopher Olah, alongside Cardinal Victor Fernández, Cardinal Michael Czerny, and theologian Anna Rowlands from Durham University. A Vatican source told the National Catholic Reporter that usually we don't invite, quote unquote, someone from the outside. But this time they did. That's our second first. And so I want you to pause here because this is unprecedented. But at the same time, I want you to think, did you catch the contradiction here? Or like the apparent paradox? Because the pope released a document that calls for, quote unquote, safeguarding the human person in the time of AI. —a document that says, literally, that technology is never neutral. And he did that standing, though, next to one of the owners of that very industry he was critiquing. So imagine if Rerum Novarum in 1891 had been released by Leo XIII on stage with Andrew Carnegie. Would you call that ethical? And I can't help thinking, isn't there a conflict of interest here? And if there is, why is nobody asking that question? Or maybe that's the wrong question. We'll further talk about this in this episode of Between You and AI. Here's your host, Andrea Iorio speaking. I'm an Italian keynote speaker and USA Today bestselling author. I was the former head of Tinder in Latin America for 5 years, chief digital officer at L'Oréal Brazil. I'm an MBA professor at Fundação Cabral and columnist at MIT Technology Review and Wired magazine. My latest book is Between You and AI, published by Wiley. You can get to know me better at andreaiorio.com and my work as a keynote speaker. So let's go back to 1891 for a moment. That's the year Leo XIII came out with Rerum Novarum, Latin for "of new things." Picture the world at that moment: factories swallowing entire villages, 12-hour shifts, children working machines that maimed adults, cities exploding in size, and two ideological forces fighting over the soul of work. On one side, unregulated capitalism. And on the other, revolutionary socialism. The Church had been silent for almost a century while the Industrial Revolution rewrote what it meant to be a human being. Then, Leo XIII broke the silence. Rerum Novarum did 3 things that nobody expected from a pope. First, it defended the right of workers to organize in unions, at a time that was considered very radical. Second, It defended private property against the socialists who wanted to abolish it. And third, it said something more important than both: that the dignity of the human person comes before capital and before the state. Neither the factory owner nor the revolutionary commissioner gets to define what a worker is worth. And so the document became the cornerstone of what we now call Catholic Social Teaching. Every encyclical on labor, economy, and human dignity for the next 135 years traces back to it. Quadragésimo Anno in 1931, written on its 40th anniversary. Centésimus Annus in 1991, on its 100th. And now, in 2026, on its 135th year, the new Pope Leo picks up the same pen. Same name, same move, different machine. Because the parallel here is almost too clean. In 1891, the question was: what does it mean to be human when the steam engine is reorganizing labor? In 2026, the question is, what does it mean to be human when the language model is reorganizing thought? So jump to May 25th, 2026. The Vatican released the first encyclical of Pope Leo XIV's pontificate, Magnificat Humanitas, and the subtitle is On Safeguarding the Human Person in the Time of Artificial Intelligence. 235 pages, 42,000 words, 5 chapters, and one central image that repeats throughout. Humanity must choose: either build a new Tower of Babel, or build the city where God and humans dwell together. Babel in the encyclical is defined this way: idolatry of profit, uniformity that erases differences, and— this is the phrase that hit me the hardest— the "reduction of human mystery to data." And let me sit with that image for a second, because it's actually very strong. Go back to the original story. Genesis chapter 11. Humanity after the flood decides to build a tower that reaches the sky, not out of curiosity, but out of ambition. And the verse is clear, they wanted to make a name for themselves. They wanted to become like God without God. And the punishment in the story isn't fire, isn't water, isn't plague. It's the confusion of languages. Suddenly nobody understands anybody. The project collapses not because the tower fell, but because the builders could no longer talk. To each other. Now hold that next to what Pope Leo is saying about AI. He's not saying AI is the new tower. He's, he's much smarter and it's deeper than that. He's saying AI risks becoming the new Babel impulse, the dream that we can engineer our way out of being human, that we can optimize away vulnerability, automate away judgment, scale away difference until every person looks like every other person and every decision looks like every other decision. A world where the marketing department in São Paulo and one in Stockholm produce identical campaigns because they're prompting the same model. A world where the candidate from Lagos and from Lisbon get filtered through the same algorithm with the same biases. A world where, as the encyclical says, the mystery of the human person gets reduced to a row in a database. And honestly, I agree with the Pope. Look, I read the encyclical and I identified 5 overlaps with chapter 10 of my own book, Between You and AI. It's a chapter that I call Agency. It's a skill that I cover. And I'm not saying the Pope copied me, God forbid. But what I want to say is that there are very, very resounding similarities. And let me go through them one by one. First is the fact that the danger isn't AI, but it's decision laundering. So the encyclical's sharpest warning is about humans who stop being authors and become kind of like forwarders of machine outputs. You know, like whenever we click and we just, you know, send out that meeting transcript. Yeah, that's what we tend to do with decisions, especially for decisions about employment, credit, access, reputation, and so on. Systems don't know compassion, mercy, or the hope that people can change. Humans do. And that's why human in the loop must mean actually in the loop, not rubber stamping. Second, agency is bilateral. Builders own it too. Pope Leo said that those who design or finance systems that thrive on human weaknesses bear a moral responsibility that cannot be ignored. And so agency doesn't only apply to the person prompting. It applies to the person shipping, not an accident. Again, that Anthropic's Chris Ola, head of AI interpretability research, was on stage next to the hope. Third, Babel is what happens when agency collapses at scale. So the encyclical's central image is a choice: rebuild Babel or build a city where humans dwell together. Babel— uniformity, idolatry of profit, the reduction of human mystery to data. It's what civilizations become when individuals stop owning their decisions. Fourth is refusal is an agency move. This is where the encyclical actually extended my own argument, because real agency includes the freedom not to use AI in certain decisions. In work, in education, in family life. And most leaders don't think of refusal as a leadership skill, but they actually should. And fifth, that agency requires selfhood. Selfhood requires limits. The encyclical deepest claim is that human limits—illness, aging, vulnerability—are not defects to be optimized away. They are how we become someone capable of taking responsibility in the first place. Without limits, no identity. Without identity, no agency. And although we find agreement on the content, I actually find myself with some doubts regarding an apparent contradiction in what took place in the Vatican launching, uh, this encyclical. As if the Pope is saying those who build the systems bear a moral responsibility that cannot be ignored, but he's saying that standing next to someone who builds these systems What's going on? Well, here there's 3 possible readings. Reading 1: The Pope got co-opted. Anthropic just got a free moral legitimation. A Vatican stamp that's worth gold in a market where every AI company is trying to differentiate itself as a safe AI company. Olaf walked onto that stage, got the papal nod, and flew back to San Francisco with the most premium ethical seat on its claim that money can't buy. Did you buy that? I didn't. I don't think it's that. Obviously, they work together. Now, reading 2 is that the Pope co-opted Anthropic. By inviting Ola specifically, and not Sam Altman, not Elon Musk, not Mark Zuckerberg, the Vatican made a strategic move. It elevated one faction of the AI industry, the safety faction, and symbolically isolated the others. So it's not an endorsement of AI. It's an endorsement of safe AI. And that's pretty different. Now, reading 3, and this is the one I find most interesting, is neither of the first 2 readings is complete. Because they both assume the Church should be neutral. And here's one thing: the Church has never been neutral. Encyclicals exist exactly for this purpose, for the Church to take sides on moral questions. Rerum Novarum wasn't neutral between capital and labor. Laudato Si' wasn't neutral on the environment. Expecting neutrality from the Pope on AI is a category error. The real paradox, then, isn't at the Vatican. It's in Silicon Valley. Think with me. AI companies sell themselves as neutral platforms. We're building the tools, the user then decides what to do with them.. But the encyclical just stated that this is all a lie. Technology is never neutral. Whoever builds technology carries responsibility. And so the right question therefore isn't whether the Pope has a conflict of interest. The right question here is why the AI industry pretends to have no moral position while the Church has the courage to admit that it does. And here we get to what I think is the real message of the encyclical and the reason why it connects directly with Chapter 10 of my book. The underlying question is one and the same: who is the author of the decision? This isn't a new question. Every technological revolution has posed the same one. In the Industrial Revolution, it was: who's responsible when a machine maims a worker? The factory owner? The worker? The state? Rerum Novarum in 1891 answered: all of them. Responsibility is distributed, but never evaporated. In the digital revolution, the same question came back as algorithms. Who's responsible when Facebook's algorithm radicalizes a teenager? The platform? The developer? The user? And the industry's answer for 20 years was: nobody. We're just a neutral platform. Think about social media. And you remember how much that answer costs us. Elections, mental health, polarization. Now, with AI, the question is back again. It's kind of like, you know, another time humanity is tripping over the same stone. Who's responsible when AI decides who gets the loan, who gets into the university, who gets fired? And the encyclical's answer is the same as 135 years ago: all of us. Whoever builds, finances, regulates, uses. Nobody escapes. My book's answer is is more operational? It's you. In front of your computer, in front of your team, in front of your decision, you remain the author. Even when AI is suggesting the answer. Even when AI is drafting the email. Even when the model tells you with 87% of confidence that that person should be let go. You are the author of that decision. So think for a moment of the last work decision you made where AI was involved. It could be an analysis, an email response, a CV screening. Now answer me honestly. Were you actually the author of that decision or you were just, you know, the forwarder? Like on a scale of 0 to 10, how much of it is yours and how much is it the machine's? If your answer is "I don't even know anymore," well, you just touched what the encyclical calls the reduction of human mystery to data. So no, I don't think the Pope has a conflict of interest for having invited Ola. I think he did something smarter. He forced both sides of the table, the church and Silicon Valley, to admit in front of each other that nobody in that room was neutral. See, 135 years ago, Pope Leo XIII wrote Rerum Novarum in response to the Industrial Revolution. Yesterday, Pope Leo XIV wrote Magnifica Humanitas in response to artificial intelligence. Now, the technology changed from the steam loom to the language model.. But the central question, though, hasn't moved an inch: human responsibility, first in what we build, then in what we do with it. Christopher Olah, the Pope, and you, the three of you are actually on the same stage every time you make a decision or perform a task using AI. And with this, I wanted to thank you so much for your attention, uh, in this episode of Between You and AI. If you liked it, forward it to a colleague, a friend, a family member that might be interested in the impact of AI on the future of work. Here's your host, Andrea Iorio. I'm a keynote speaker and USA Today bestselling author. You can get to know more about me on my website, andreaiorio.com. Thank you so much, and see you, uh, in the next episode of Between You and AI.