Summary
- Pope Leo XIV introduced "Magnifica Humanitas" on May 25, marking the first encyclical focused solely on artificial intelligence.
- The document asserts that algorithms, data, and digital platforms should be recognized as common goods, not monopolized by private entities.
- Christopher Olah, co-founder of Anthropic, warned during the Vatican's launch that the potential for widespread AI-induced job displacement presents a "historic moral imperative" that must be addressed.
Pope Leo XIV unveiled his inaugural encyclical on Monday, a comprehensive 245-paragraph document centered on artificial intelligence, which calls for stricter regulation of Big Tech, characterizes data as a collective human resource, and posits that "technology is never neutral," as it reflects the values and biases of its creators.
The encyclical, titled Magnifica Humanitas ("Magnificent Humanity"), was presented at the Vatican's Synod Hall on May 25. Pope Leo signed it on May 15, coinciding with the 135th anniversary of Rerum Novarum, an encyclical by Pope Leo XIII addressing labor rights and laying the groundwork for contemporary Catholic social teaching.
Pope Leo has consistently highlighted AI as a pivotal ethical challenge of his papacy, likening the anticipated societal shifts to those experienced during the Industrial Revolution.
The encyclical addresses a wide array of topics, including the use of AI in warfare, dehumanization, technocracy, data exploitation, child safety online, mass unemployment, misinformation, autonomous weapons, and transhumanism. However, its core argument is straightforward: every algorithm embodies the priorities of its creators. Constructing systems that deny this reality does not eliminate bias; it merely conceals it.
Data is a shared asset, belonging to everyone, including individual users.
Traditional Catholic teachings maintain that the earth's resources are meant for all, not just for private ownership. Pope Leo extends this principle to the digital realm, asserting that algorithms, platforms, and data should be managed as common goods rather than being confined by a few corporations.
"Data is generated by numerous contributors and should not be commodified or entrusted to a select few," the pope stated.
The encyclical also emphasizes the principle of subsidiarity, which advocates for decisions to be made at the most local level possible, specifically in the context of tech platforms. It calls for not only top-down regulations but also for transparent algorithms, independent community audits, and substantial legal recourse for individuals affected by automated systems impacting their credit scores, job applications, or legal risk assessments. According to Leo, without this distributed oversight, AI governance could evolve into a form of digital authoritarianism that stifles the voices of those it claims to serve.
The encyclical critiques transhumanism, the notion that human limitations should be eliminated. Leo argues that human vulnerability is not a flaw, but rather a foundation for empathy, moral judgment, and genuine care for others. Systems designed to eradicate these qualities do not foster better humanity; they create mechanisms that more effectively marginalize the vulnerable.
Importantly, the pope refrains from attributing human-like qualities to technology. The encyclical states that AI systems "lack a body, do not experience joy or pain." It emphasizes that these systems do not possess the lived experiences necessary for true understanding, as they can simulate empathy and generate convincing language, yet lack comprehension of their output.
This distinction has practical implications. When an algorithm makes decisions regarding hiring, sets credit terms, or determines risk scores in legal situations, its perceived objectivity can obscure the inherent biases introduced by its designers. The encyclical cautions against delegating critical decisions to automated systems that "lack compassion, mercy, and forgiveness" and warns against interpreting such outputs as neutral merely because they are machine-generated.
During the event, Christopher Olah, co-founder of Anthropic and head of its interpretability research team, shared the stage with Pope Leo, drawing significant attention. Olah openly acknowledged that major AI labs "operate within a framework of incentives and constraints that sometimes conflict with ethical conduct," emphasizing that external scrutiny from governments, religious bodies, and civil society is essential. He also highlighted the imminent threat of AI-related job displacement, which, if it occurs on a large scale, would create "a moral imperative of historic proportions."
Pope Leo has articulated a more challenging version of this argument. He states in the encyclical that "a more moral AI is insufficient" if its ethical framework is determined solely by those who control the data and computing resources. Earlier in November 2025, he directly conveyed this message to Silicon Valley leaders at the Vatican. Additionally, on May 16, the Vatican established a new internal AI commission, drawing from seven departments to coordinate efforts on AI governance moving forward.