Olah and Pope Leo XIV press AI power limits

Olah and – Anthropic cofounder Christopher Olah appeared alongside Pope Leo XIV to help unveil an AI-focused encyclical at the Vatican—arguing that AI’s risks grow when power is concentrated and insisting that moral and religious dialogue must stay in the room.

By the time the Vatican unveiled its latest teaching on artificial intelligence, the surprise wasn’t just that Pope Leo XIV had reached a first: becoming the first pope to personally present an encyclical to his followers. The harder-to-ignore detail was who stood alongside him.

Anthropic cofounder and self-proclaimed atheist Christopher Olah was there too. helping unveil the encyclical titled “Magnifica humanitas: on safeguarding the human person in the time of artificial intelligence.” The Vatican doesn’t usually pull tech outsiders into the spotlight—especially not figures who build systems at the center of today’s AI boom. But Leo’s message carried a sense of urgency, grounded in what he called “the gravity of the moment.”.

In the encyclical—and in the presentation—Leo framed the moment as one where a few powerful people control the fate of the world. He didn’t name individuals. Still, his warnings landed with particular force on the billionaire executives who steer the largest tech companies.

“When such power is concentrated in the hands of a few,” Leo warned in the letter, “it tends to become opaque and evade public oversight, increasing the risk of distorted forms of development that give rise to new dependencies, exclusions, manipulations and inequalities.”

He went further, calling for “freeing technology from monopolistic control and opening it to discussion and debate” by “making it human-friendly and restoring it to the plurality of human cultures and ways of life.”

Olah, an unlikely companion in a Vatican setting, didn’t challenge the thrust of Leo’s argument. If anything, he sounded like someone trying to keep AI ambitions tethered to human limits.

Out of the gate. Olah presented himself as a different kind of tech founder—one who is cautious. even worried. about AI. He began his speech by calling himself a person who got into AI work because he “has a desire to help things go well for humanity.” But he also warned that good intentions don’t inoculate anyone against getting pulled into the logic of capability.

“Every frontier AI lab, including Anthropic, operates inside a set of incentives and constraints that can sometimes conflict with doing the right thing,” he told the audience.

Olah also echoed Leo’s central concern: danger increases when AI is controlled by a powerful few. He said: “AI development is concentrated in a handful of wealthy nations. How can we ​ensure the gains of AI are shared globally?”

For Olah, the stakes weren’t only about geopolitics or fairness. He returned to an issue many people feel in their daily lives—the fear of losing work. He warned that AI could “displace human labor,” adding that there is “a real possibility that AI will displace human labor at a very large scale.”

“If that happens, supporting those who have been displaced will be a moral imperative of historic proportions,” he said.

That combination—AI leadership with an emphasis on interpretability research and human control—was part of why his presence in the Vatican matters. Olah is known for interpretability, or making machines more human-friendly. He previously led interpretability research at OpenAI. and his insistence that humans must remain in control of the technology lines up closely with the pope’s concern about oversight and dependence.

As the speeches continued, Leo and Olah converged again on a point that felt less technical, but more demanding: they both called for continued collaboration between the people building AI and the people tasked with judging its moral direction.

Olah argued that even if scientists and mathematicians are the ones creating powerful technology, they aren’t automatically the ones best equipped to decide how it is used. He urged conversations between tech leaders and religious and moral leaders to keep AI in check.

“It is through dialogue and mutual effort, through the push and pull, that humanity will achieve great things,” Olah said. He added: “That is what I see in Magnifica Humanitas, and it is why I am grateful to his holiness and to the church for taking up this work of discernment.”

Leo answered with his own version of the same invitation. He said, “The church wishes, with humility and frankness, to be part of conversations on artificial intelligence.” He explained that the church may not have “technical answers,” and that it does not want to “displace those with expertise.”

What it offers instead, Leo said, is “a wisdom concerning the human that our present time desperately needs.”

Taken together, the event placed a blunt question in the center of the AI conversation: when power sits with a small set of powerful actors, who gets to oversee what comes next—especially when the technology could reshape labor, deepen inequality, and create new forms of dependence.

At the Vatican, at least, the people holding the microphones agreed that the answer should not come from the powerful alone.

Pope Leo XIV Christopher Olah Anthropic Vatican encyclical artificial intelligence AI oversight interpretability labor displacement technology monopolies human-friendly AI