In a surprising twist, an independent analysis indicates that parts of Pope Leo XIV’s latest encyclical, Magnifica Humanitas, may have been influenced by artificial intelligence. The document, which addresses the ethical implications of AI on humanity, was released earlier this month and has sparked discussions about the intersection of technology and traditional institutions.
AI detection tools flag suspicious patterns in encyclical
A detailed examination by researcher Linch Zhang, published on the LessWrong forum, highlights that certain paragraphs in the encyclical share characteristics commonly associated with AI-generated text. Zhang’s analysis, which leverages the Pangram AI detection tool, estimates that between 40% and 100% of some sections were likely composed by AI. These patterns include an unusually high frequency of words like "genuinely," a term frequently used in outputs from Anthropic’s Claude model.
Zhang’s findings were further corroborated by another LessWrong user who conducted a section-by-section evaluation of the document using Pangram. Their analysis suggested that 62% of the first chapter alone contained AI-like phrasing, reinforcing concerns about the encyclical’s origins.
Theological implications of AI-assisted writing
The possibility that AI contributed to an official Vatican document raises ethical and theological questions. Encyclicals are traditionally authored by the Pope or high-ranking clergy, serving as authoritative teachings for the global Catholic community. If AI played a role in drafting sections of Magnifica Humanitas, it could challenge long-standing norms around authenticity and human authorship in religious texts.
Some observers argue that AI’s involvement could democratize content creation, enabling institutions to produce documents more efficiently. Others caution that reliance on AI risks diluting the personal and spiritual intent behind these communications. The Vatican has not yet issued an official response regarding the analysis or the potential use of AI in crafting the encyclical.
What the tools reveal about AI-generated text
AI detection tools like Pangram analyze writing for statistical patterns that are statistically improbable in human-authored text. These include repetitive phrasing, an overuse of certain adverbs, and a lack of idiosyncratic stylistic flourishes. In the case of Magnifica Humanitas, the detected anomalies align with patterns observed in outputs from major language models.
While AI can generate coherent and contextually relevant text, its limitations remain evident. For instance, the tool’s analysis suggests that the AI-generated sections lacked the nuanced theological reasoning typically found in encyclicals. This discrepancy underscores the current challenges in using AI for complex, value-driven content.
Looking ahead: AI’s role in institutional communication
The findings regarding Magnifica Humanitas may prompt institutions—religious, academic, or governmental—to reconsider how they integrate AI into their communication strategies. As AI tools become more sophisticated, distinguishing between human and machine-generated content could grow increasingly difficult.
For now, the Vatican’s silence on the matter leaves room for speculation. Whether AI played a larger role in the encyclical’s composition remains unconfirmed, but the discussion itself highlights the growing need for transparency and ethical guidelines around AI’s use in influential documents.
AI summary
Papa Leo XIV’ün yeni encyklikasında AI izleri bulundu. Yapay zekânın dini metinlere etkisi ve belgenin hazırlanma sürecinde AI kullanımına dair ayrıntılar.
