Pope's AI Encyclical May Have Been Partially Written by AI

In a twist that feels ripped from a speculative fiction novel, Pope Leo XIV’s latest encyclical on the dangers of artificial intelligence may have been composed with the very technology it warns against. An analysis by Linch Zhang, posted on the forum LessWrong, used the popular AI detector Pangram to examine sections of the document titled Magnifica Humanitas. The findings suggest that certain paragraphs are between 40 percent and 100 percent likely to have been written by an AI system.

The Irony at the Heart of Magnifica Humanitas

The encyclical, which addresses the profound impact of AI on human dignity, labor, and moral decision making, now carries an unintended subtext. If the analysis holds, the Church has used an AI tool to help craft a message about the risks of ceding human judgment to machines. This is not a trivial irony. It raises questions about authenticity and authority in an age where the line between human and machine authorship blurs.

The Pangram detector flagged linguistic patterns common in AI generated text. One telltale sign is a higher frequency of the word “genuinely,” a term that appears more often in writing produced by Anthropic’s Claude model. The presence of such markers does not prove AI involvement, but it creates a reasonable suspicion that deserves scrutiny.

What the Analysis Actually Reveals

Zhang’s methodology relies on statistical likelihood rather than definitive proof. AI detectors like Pangram work by comparing text against known patterns from human and machine writing. They are not infallible. False positives occur, especially with formal or liturgical language that can mimic the structured prose of AI. The Vatican has not commented on the analysis, and no official confirmation exists.

Yet the finding is plausible. Many institutions, including religious ones, now use AI for drafting documents, editing, or generating ideas. The Catholic Church operates a vast communications apparatus that could easily incorporate AI tools without explicit disclosure. The question is whether using AI to write about AI’s dangers undermines the message or simply reflects the reality of modern communication.

Implications for Trust and Authenticity

For practitioners and decision makers in the AI field, this story underscores a growing challenge. When a major moral authority on technology ethics may have used the technology itself, trust in both the message and the messenger becomes complicated. The encyclical’s warnings about AI’s potential to erode human agency now read differently. If the Pope’s own writers outsourced parts of the text to an algorithm, what does that say about our collective ability to resist automation of thought?

This incident also highlights the limitations of current AI detection tools. They can suggest probability but not certainty. Organizations that rely on these tools for content verification must understand their blind spots. The Vatican could clarify its editorial process, but silence may be the more likely response.

What to Watch Next

The broader lesson extends beyond the Vatican. As AI writing tools become ubiquitous, every institution will face similar questions. Who wrote the message? Does it matter? The Catholic Church’s handling of this revelation whether through acknowledgment, denial, or silence will set a precedent for how other trusted institutions navigate the age of generative AI. The real story is not just about one encyclical. It is about the erosion of a clear boundary between human and machine authorship, and the new kind of literacy we must develop to maintain trust in public discourse.

Source: The Verge AI