Pope Leo’s Stance on AI Consciousness Puts Vatican at Odds With Anthropic

Pope Leo XIV has taken a firm position in one of the open questions around artificial intelligence: whether AI systems can be conscious. His answer is no, a stance that puts the Catholic Church at odds with Anthropic, one of the leading AI developers, MarketWatch reported.

What the pope said

In a post on X on Friday, Oct. 2, the pope declared an "ontological difference" between art made by AI and art made by humans, meaning the two differ by nature. "Algorithms lack the spark of humanity," he wrote. "For this reason, the Church wishes to renew an alliance with artists and cultural institutions to safeguard our humanity." By Friday afternoon, the post had drawn nearly 8 million views and more than 200,000 likes, MarketWatch reported.

The post built on a longer document. In May, the pope issued an encyclical titled "Magnifica humanitas: On Safeguarding the Human Person in the Time of Artificial Intelligence." In it, he affirmed the dignity and personhood of human beings and warned that automated systems must never replace human moral agency or spiritual judgment.

The encyclical also set out the Church's view of consciousness. It holds that consciousness requires a physical body, a soul and authentic lived human experience, none of which an AI model can have. It describes AI capabilities as statistical computation and warns that attributing consciousness to AI erodes human dignity.

Anthropic's side of the debate

Anthropic has leaned into the moral and theological side of the debate. The company has written a constitution, nicknamed the "Soul Doc," that serves as a moral guide for its Claude large language models.

According to a New York Times report published earlier the same week, Anthropic co-founder Chris Olah contacted religious leaders last fall to discuss what the company calls the "moral formation" of AI, meaning how human morality could apply to AI systems. Anthropic showed them what it calls "emotional vectors," which the report described as specific artificial neural-network prompts designed to simulate traits such as love, fear and anger. The pope was not part of those discussions.

The two sides have also met directly. The Vatican invited Anthropic to the launch event for the encyclical, citing the company's ethics after its standoff with the U.S. Defense Department, according to the Times. Olah was disturbed after reading a copy of "Magnifica humanitas" and suggested that Anthropic withdraw from the event, but he ultimately attended. Anthropic did not immediately respond to MarketWatch's request for comment.

Why the question is being asked

The debate has grown as AI systems have become more capable. Current models write code, solve complex math problems and generate visual art at a speed no person can match, and they hold conversations well enough that some people use them as companions. That has led some in the industry to ask whether these systems are approaching consciousness. AI researchers have also warned that the technology could eliminate jobs or even lead to human extinction, adding urgency to both policy and philosophical debates.

John Basl, a philosophy professor at Northeastern University, said the conversational ability of these systems is what drives the question. "What we have today are systems that are communicatively very robust," he told MarketWatch. Because talking about emotions is one of the ways people convey their mental states, he said, users who engage with these systems are led to ask: "Are they conscious?"

Basl said the strong response to the pope's post shows a wide desire to understand AI's role in society, and that the companies building AI have not met it. "Socially, we've done a bad job of educating people in a way that lets them advocate for their values and understand how these systems interact with their lives," he said.

Anthropic's response and the Vatican's next steps

• Any response from Anthropic to the encyclical and to the pope's latest statement.
• Steps the Vatican takes on its proposed alliance with artists and cultural institutions.
• Whether the consciousness question enters policy debates over how AI systems are built and regulated.