“It is enormously important that there be people outside those incentives,” Olah said, describing the need for “earnest, thoughtful, critics” willing to challenge the industry when necessary.
The appearance marked a rare public alignment between a major AI company and the Catholic Church on the need for broader oversight of advanced AI systems. Rather than framing the issue as a purely technical challenge, both the Vatican and Olah emphasized that decisions around AI development increasingly involve questions traditionally associated with philosophy, ethics, and religion.
Olah told attendees that modern AI systems differ fundamentally from conventional engineered technologies because researchers do not fully understand how their internal behaviors emerge. He described the models as “grown” from vast amounts of human language and thought rather than constructed piece by piece like traditional machines.
“We keep finding things that are mysterious, even unsettling,” Olah said, pointing to internal model behaviors that researchers believe resemble forms of introspection and emotional states.
Leo XIV’s encyclical, whose Latin title translates to “Magnificent Humanity,” centers on protecting human dignity as AI systems expand into everyday life. The document raises concerns about autonomous weapons, biased algorithms, and the concentration of AI development within a small number of wealthy nations and corporations.
The pope argued that technology itself is morally neutral, but said societies must decide whether AI will reinforce systems driven primarily by self-interest or support communities centered on human well-being. He also warned that the financial incentives surrounding AI make industry self-regulation insufficient on its own.
“The church wishes, with humility and frankness, to be part of conversations on artificial intelligence,” Leo XIV said during the presentation. “We do not possess technical answers, nor do we seek to displace those with expertise, but we bring a wisdom concerning the human that our present time desperately needs.”
A major focus of the discussion involved the potential economic consequences of AI deployment. Olah warned that large-scale labor displacement could become a defining moral challenge for governments and institutions worldwide, particularly because the gains from AI development are currently concentrated in a small number of countries.
“How can we ensure the gains of AI are shared globally?” Olah asked. “We do not have a mechanism for this.”
The Vatican also used the event to push for broader international cooperation on AI governance. Leo XIV urged religious leaders and governments to work together to establish ethical guardrails before the technology becomes more deeply embedded across society.
“I ask you to show the world that we are united in asking for a proactive commitment to protect human dignity in this new era of machines,” the pope said in a message to attendees at a related AI ethics conference involving representatives from multiple faith traditions.
Senior Vatican official Fr. Brendan McGuire described the effort as urgent, arguing that institutions hoping to influence the direction of AI development must engage now rather than after the technology becomes fully entrenched.
“What they have asked for is partnership, and it would be morally reprehensible for us to not partner with them,” McGuire said. “This technology is developing, and if we want to influence it, we need to do so now.”
This analysis is based on reporting from anthropic.
Image courtesy of Anthropic.
This article was generated with AI assistance and reviewed for accuracy and quality.
