The argument boil down to: Pangram seems to flag some sections, some sections have tells or AI smell, and despite being written in a similar register, previous writings of both the Pope and previous encyclicals are not flagged.
I'm not sure that it's a disaster if parts have been AI assisted, but it's an interesting question to ponder, and would be interested in the HN view.
Best post I've found arguing that there was LLM use: https://substack.com/inbox/post/199316613
Australian startup apparently verifying (without much detail) that it's not: https://www.proudlyhuman.org/news-room/media-releases/pope-l...