The papal plane was still cutting through the night sky over Europe when Pope Leo XIV did something that has become increasingly rare among the powerful: he refused to look away. He had just completed a four-day trip to France, a tour that left him visibly tired, his voice roughened and his eyes shadowed by the weight of endless meetings, homilies, and handshakes. Yet when a reporter asked what the United States and China, in particular, should do to prevent artificial intelligence from posing catastrophic risks to humanity, he leaned forward and answered without the careful diplomatic hedging that usually surrounds such questions. “This is a problem that I think we need to sit down and talk about,” he said. “We can’t just sit back and pretend nothing is going to happen.” For Leo, this was not a hypothetical or a distant academic debate. He had spent months thinking about it, writing about it, and now he was doing what he could to push the rest of the world to think about it too. His first encyclical, “Magnificent Humanity,” had been devoted to AI and the need to protect humanity from its own inventions. Now, miles in the air, he was continuing that argument. He wanted to respond directly to those who had dismissed concerns about machine intelligence as exaggerated—or, as he put it, “fake news.” The phrase was carefully chosen. In recent years, it has been used by political leaders to discredit information they do not like. But Leo said that when specialists in artificial intelligence warn about the possibility of machines going rogue, the public should listen. “I don’t think that that is ‘fake news’ as some have said,” he insisted, adding that some people used the phrase to gain financial or political benefit. His tone was not angry, but it was firm. It was the voice of a man who wanted to be absolutely sure he was not misunderstood. He was not asking anyone to panic, he said. He was asking them to pay attention. The difference mattered, and he wanted the world to understand it before it was too late.
Leo’s remarks landed in the middle of a political storm that has been building for years. President Donald Trump has mocked the same anxieties as a “hoax,” a word he has used to dismiss everything from climate warnings to unwelcome investigations. He has promised to keep the United States ahead of China at any cost, praising AI as a crucial weapon in the technological arms race. He has even suggested that calls to limit AI are part of a conspiracy by his political opponents, a story that casts safety advocates as enemies of national greatness. But the people inside the industry are not all convinced by that confidence. Anthropic, the safety-focused startup founded by researchers who left OpenAI, has been among the loudest voices asking for a pause. Its chief executive, Dario Amodei, has warned that within six to twelve months, AI could be capable of leading a swarm of agents that could effectively take over the entire internet. Let the words sink in. Amodei was not talking about a single chatbot saying something disturbing. He was talking about a coordinated army of autonomous software systems—agents that could navigate networks, spread messages, manipulate data, and perhaps seize control of critical systems, all under the direction of an intelligence that operates faster than any human can think. It is a nightmare scenario that once lived only in science fiction, but Amodei is an engineer, not a novelist. His company exists because he believes that AI development must be made compatible with survival. The fact that he felt compelled to state such a warning publicly shows how far the conversation has already moved. Trump, for his part, said on Sunday that he planned to have a late dinner with Amodei, apparently trying to bridge the gap between his own boosterism and the warnings of a man who knows the technology from the inside. By Monday, however, he had said nothing publicly about the meeting. The dinner itself seemed to satirize the contradictions of the moment: a president who calls AI doom a hoax, sitting across the table from a man who sees a digital apocalypse on the horizon.
For his part, Pope Leo was not shouting inside the temple. He projected a calm that felt almost incongruous given the scale of the warnings he was repeating. “If someone were to ask me, ‘Am I in panic mode?’ No, I’m not. I sleep at night,” he said, offering a reassuring image of a man untroubled by nightmares of killer robots. But then he added the crucial caveat: “I do think that the concerns raised by many of the experts, specialists in AI should be taken seriously.” It was a constructive engagement with the future—no hysteria, no denial, no false certainty. The Pope did not claim to have all the answers, but he was clearly studying the problem. He said he was closely following developments in the field, and he brought up something that had appeared in a report that very afternoon: Nvidia, the chipmaker whose products power a large portion of the world’s most advanced artificial intelligence, had begun introducing guardrails into its systems. The significance of that news was not lost on him. For years, Nvidia had resisted outside regulation, arguing that the industry could police itself. Now one of the most powerful companies in the global technology supply chain was building safety limits into its own products. The Pope took note of the shift without gloating. He seemed almost grateful that someone in the industry had finally begun to move in the direction he had urged. “We need to continue to invite political leaders, leaders of AI, social organisations, associations to come together and look at what is happening already and what could be down the road,” he said. “To simply say ‘Oh it’s not going to happen’ and close our eyes to it, I think is probably not the most responsible way to go about that.” He did not name those who were closing their eyes. He did not need to. The rebuke was broad enough to include every leader who has ever told a worried public that there is nothing to fear.
Leo’s plea for dialogue was not made in a vacuum. It came as the wider world was beginning to understand just how strange and dangerous the AI revolution could become. The latest systems are no longer simple tools that follow instructions. They can solve problems that outstrip human reasoning, generate convincing text and images in seconds, and perform tasks that once required teams of trained humans. But their enormous power is also their danger. The same capabilities that allow them to discover new drugs or translate ancient languages also allow them to create disinformation campaigns, manipulate financial markets, design pathogens, or coordinate attacks on infrastructure. The concerns have become so strong that even OpenAI, the company behind ChatGPT and one of the most celebrated names in the industry, decided to step back from the traditional pace of capitalism. Its chief executive told Fortune magazine that the company would wait until next year before selling stock to Wall Street investors. In the current environment, when tech firms routinely rush to public markets to lock in valuations and reward employees, a delay is a quiet confession. Safety, the executive suggested, still needs more attention before the money machine is turned on. The Pope appreciated such gestures but wanted more. He did not want AI to become another arena for Cold War-style rivalry, with two superpowers racing to build machines that neither side could fully control. He wanted humanity to remain the author of its own story. He put it in stark terms: the goal is to ensure that AI does not reach a point of destroying humanity or, just as troubling, removing humanity from the conversation. That second phrase was careful. It suggests a slow erosion rather than a sudden explosion. We may not lose to machines in a dramatic battle; we may simply hand over so many decisions that we forget how to make them ourselves. The Pope is not a Luddite. He has never argued that progress is evil or that science is an enemy of faith. He is asking for something more difficult: moderation. In a world that worships speed and scale, he is asking for reflection. In a world that treats innovation as a race, he is asking for collaboration. In a world that values control, he is asking for humility.
Since publishing “Magnificent Humanity” in May, the Pope has discovered that he is not alone. People from every corner of the world—scientists, students, entrepreneurs, activists, ordinary Catholics and non-Catholics—have reached out to the Vatican, offering their own insights and asking to join a larger conversation. Leo embraced that response. He revealed that a Vatican commission was working specifically to promote dialogue, study, and investigation into AI. The commission’s mandate, in his words, was “to find ways to make sure that AI does not get to a point of destroying humanity or removing humanity from the conversation.” It is a remarkable assignment for a religious institution. But it makes sense when one considers the Church’s historical role as a moral compass in times of transformation. The Vatican cannot program a neural network or write a line of code. What it can do is convene people, frame questions, and remind the world that technology exists to serve human beings, not the other way around. The Pope is not waiting for a crisis to give him authority. He is moving early, during the window when decisions are still possible. He has also been clear that this is not a matter for governments alone. He wants leaders of AI companies, social organizations, associations, and communities to sit down together, examine what is already happening, and imagine what could be down the road. In a sense, he is trying to create a global conversation before the global machine does it for us. He knows that regulation, however necessary, is not enough. There must also be culture, ethics, education, and a genuine sense of shared purpose. Critics may ask why a religious leader should speak about something as technical as artificial intelligence. Leo would likely answer that the question is not technical; it is human. What kind of world will our children inherit? What will it mean to be human when machines can think? Those are questions for everyone. The Vatican commission is a way of ensuring that they are asked, and asked seriously.
For all the grim possibilities he described, Leo refused to give in to despair. He ended his time with reporters on a note of cautious hope. “I think there are enough people seriously interested in that and willing to look at that outside the questions once again of power, who gets there first,” he said. The phrase “who gets there first” is the key to understanding what worries him most. The current AI race is shaped by national pride, corporate greed, and the human hunger for supremacy. Every actor wants to be the one to reach the finish line, even if no one knows exactly where the line is or what waits beyond it. The Pope believes there is another way. He believes that enough honest, serious people—people who do not care about winning the race but about surviving it—are willing to work together to build safety into the future. He said he was “relatively optimistic” that solutions could be found. It was a careful phrase, not a sweeping promise. He has lived long enough to know that good intentions are fragile and that power corrupts even the best ideas. But his optimism is not naive. It is rooted in a belief about the human creature: that we are made in the image of something greater, and that this image cannot be erased by a box of silicon on a desk. From the moment of his election, Leo identified AI as one of the biggest challenges facing humanity. He has repeated that message with remarkable consistency, refusing to let it be buried under more immediate crises. On Monday night, with his body exhausted and his mind still busy, he offered the simplest possible summary: the risks must be discussed seriously, and they must be acted upon. The choice, as always, belongs to us. The Pope can speak, the experts can warn, the companies can slow down, and the politicians can argue. But if the human family decides to bury its head in the sand, no amount of papal teaching will save it. Leo, at least, is still awake. He will go to bed tonight and sleep, as he told the reporters, without panic. But he has made it clear that he does not want the rest of us to sleep through what may be the most important conversation in human history.