Artificial intelligence could displace human jobs “at a very large scale” if left unchecked, Anthropic co-founder Chris Olah warned during a Vatican event centred around Pope Leo XIV’s first encyclical on AI and ethics.

Speaking at the Vatican on Monday, Olah described the current moment in AI development as both transformative and deeply uncertain, saying governments, institutions, and society at large needed to engage more actively with the technology rather than leaving its direction solely to Silicon Valley companies.

“Every frontier AI lab, including Anthropic, operates inside a set of incentives and constraints that can sometimes conflict with doing the right thing,” Olah said. "There is a real possibility that AI will displace human labor at a very large scale,” Olah warned. “If that happens, supporting those displaced will be a moral imperative of historic proportions," he further continued.

Olah, who co-founded AI company Anthropic and is known for his work on understanding how advanced AI systems function internally, said concerns around the technology were no longer limited to engineering questions. “The ethical questions raised by AI extend far beyond engineering,” he said, welcoming the Catholic Church’s decision to engage directly with the issue. He also said public anxiety around AI, particularly among younger people, was understandable given the pace at which the technology is evolving.

Among the biggest concerns, Olah pointed to the possibility of widespread labour disruption as companies increasingly automate white-collar and knowledge-based work using generative AI systems. “There’s a very serious possibility that AI could automate many jobs at a very large scale,” he said, adding that governments and societies needed to begin thinking seriously about how economies would adapt if productivity gains from AI were concentrated among a small number of firms.

Olah also raised concerns over the unequal distribution of AI development globally. “AI development is concentrated in a handful of wealthy nations. How can we ensure the gains of AI are shared globally?” he asked during the Vatican discussion.

Another unresolved challenge, according to Olah, is that even researchers building advanced AI systems often struggle to fully understand how these models arrive at decisions. Calling it mysterious, he said that the researchers are finding it difficult to understand how these AI systems work. “There’s this question of interpretability. Can we actually understand what these systems are doing internally?” he said. "I will be honest, we keep finding things that are mysterious and even unsettling."

What is Magnifica Humanitas? 

The comments came during the presentation of Magnifica Humanitas, Pope Leo XIV’s first major encyclical since becoming pontiff. The document calls for stronger oversight of artificial intelligence and warns against the unchecked concentration of technological power. The encyclical also highlights concerns around misinformation, autonomous weapons, labour displacement, and the widening gap between countries leading AI development and those being left behind.

This encyclical is happening after 135 years, after his namesake Pope Leo XIII published Rerum Novarum, which was the foundational 1891 encyclical that happened the Industrial Revolution, focusing on labour rights.