Did the Pope Use AI to Write About the Dangers of AI?

The Papal Encyclical and AI: A Blend of Tradition and Technology?

Did the Pope use AI to write about the dangers of AI? That’s the question that has been swirling around after an analysis suggested parts of his latest encyclical, “Magnifica Humanitas,” might have been penned by artificial intelligence.

A Skeptical Look at the Papal Encyclical

It’s no secret that Pope Leo XIV is navigating a complex landscape as he addresses the ethical implications of AI. An analysis by Linch Zhang, posted on LessWrong, revealed certain paragraphs to be between 40% and 100% written by AI according to the AI detector Pangram. This isn’t just a curious incident; it’s a significant development in how religious institutions engage with cutting-edge technology.

For those unfamiliar, Pangram is an AI detection tool that has been gaining respect among researchers for its accuracy. Zhang’s findings suggest that parts of the encyclical exhibit traits common to AI-generated writing—such as an unusual frequency of the word “genuinely,” reminiscent of Anthropic’s Claude.

AI Detection Isn’t Foolproof

While these results are alarming, it’s important to note that AI detection isn’t foolproof. Different detectors can yield different results, and there’s always a margin of error. According to Pangram, its false positive rate for reporting human-written work as AI-generated is approximately 1 in 10,000.

Even so, the findings raise questions about how much of the document was indeed authored by Pope Leo XIV versus his team.

The first chapter of “Magnifica Humanitas” flagged 62% of its content as AI-generated, while a transcript of the Pope’s speech was rated 100% human.

It’s worth noting that other sections were deemed less AI-influenced, with the first 20 paragraphs of the last four encyclicals and a speech by Pope Leo XIV itself rated as 100% human. This suggests a mix of human and AI contributions in different parts of the document.

The Context: A First for the Vatican

“Magnifica Humanitas” is not only the first encyclical focusing on AI, but it also marks the first time the Pope has presented his thoughts alongside Christopher Olah, a co-founder of Anthropic. This collaboration highlights the growing importance of technology in religious discourse.

Given the Vatican’s traditional stance on matters of faith and morality, this move signals an openness to engaging with modern technologies that could influence human society profoundly.

A Wider Implication for Religious Institutions

The Pope’s use—or misuse—of AI raises broader questions about how religious institutions can navigate the complex terrain of technology. Will other religious leaders follow suit, or will they continue to rely solely on human voices? This episode is likely to prompt a deeper discussion within the Vatican and beyond.

You might think this is new, but… it’s not.

The integration of AI in religious discourse has been quietly happening for years. What’s different now is that we’re seeing it more explicitly acknowledged and debated by those at the highest levels of authority.