Analysis Flags Up to 100% of Pope Leo XIV's AI Encyclical Sections as Machine-Written
Updated
Updated · The Verge · May 27
Analysis Flags Up to 100% of Pope Leo XIV's AI Encyclical Sections as Machine-Written
2 articles · Updated · The Verge · May 27
40% to 100% of some paragraphs in Pope Leo XIV’s Magnifica Humanitas were flagged as AI-written in a LessWrong analysis by Linch Zhang using the Pangram detector.
62% of the first chapter was marked AI-generated in a separate section-by-section check, while The Verge’s test of roughly 2,000 words estimated 46% AI authorship.
Pangram also found some passages were essentially 0% AI, and the first 20 paragraphs of the last four encyclicals plus a Leo speech transcript were rated 100% human-written.
Pangram says its human-text false positive rate is about 1 in 10,000, but the report notes AI detection remains imperfect and different tools can produce conflicting results.
The Vatican did not immediately comment on the pope’s first encyclical, a document focused on AI that was presented alongside Anthropic co-founder Christopher Olah.
Did the Pope use an AI to write his powerful warning against artificial intelligence?
As AI blurs authorship, can we still trust the authenticity of our leaders' words?
Pope Leo XIV’s *Magnifica Humanitas*: 245 Paragraphs on AI, Human Dignity, and the Vatican’s New Role in Tech Ethics
Overview
Pope Leo XIV’s encyclical, Magnifica Humanitas, published on May 25, 2026, quickly gained extensive media attention for its moral and ethical guidance on artificial intelligence. Major news outlets consistently attribute the document solely to Pope Leo XIV, with no public controversy or evidence suggesting AI involvement in its drafting. The encyclical is widely discussed for its focus on human dignity and ethical responsibility in the digital age, and there are no reports from AI detection tools or technology platforms questioning its authenticity. This clear attribution reinforces the Pope’s role as a leading moral voice on technology’s impact.