I finally got around to reading Pope Leo XIV’s new AI encyclical, Magnifica Humanitas, and suprisingly to me, I think it is one of the better things I’ve read on AI ethics in quite a while.

What I found interesting is that it barely really talks about AI as software. Instead, it frames AI as part of the next Industrial Revolution and as infrastructure that will shape labour, economics, institutions and human decision making for decades to come. It's not a coincidence, Pope Leo XIV was named in honour of Pope Leo XIII, who is known for the 1891 encyclical Rerum Novarum on the rights and duties of capital and labour during the Industrial Revolution. The Pope's thinking is much less interested in apps, products or hype cycles and much more interested in the systems that form around technology and the effect those systems have on people.

Personally, As my own journey with AI continues to develop, I think that is a far more useful way to think about AI than most of the current conversation. A lot of AI discussion still lurches between productivity gains on one side and AGI or existential risk on the other. The encyclical spends very little time on either. Instead, it focuses on concentration of power, dependency, labour, automated decision making and the assumption that optimisation and productivity automatically equal progress.

That is important to me because AI is very quickly (perhaps too quickly) becoming part of the infrastructure through which society operates. Recommendation systems shape what we see, ranking systems shape visibility and opportunity, automated workflows increasingly sit inside institutions, policy and law is being digitised and large language models are starting to become the interface layer between people and information. We often talk about AI as though it is a tool sitting on top of the world, but it is becoming the operating environment itself. When you stop and think - that needs much deeper consideration than perhaps it is being given.

The document also spends a lot of time talking about human judgement, dignity and compassion. Whether you are religious or not, I think the practical questions underneath that are pretty relevant. What happens when institutions begin handing more and more judgement over to optimisation systems? What gets lost when efficiency becomes the dominant way we measure success? One passage argues that "human dignity cannot be reduced to data points, behavioural predictions or economic output," another warns that "compassion is diminished whenever suffering is treated as a logistical problem rather than a human encounter." It's really powerful stuff. The encyclical repeatedly restates the idea that good societies depend not just on accurate systems, but on people capable of empathy, moral responsibility and care for one another.

We all know as demostrated through social media and current world events that this is not the way the world seems to be going. Also many organisations are rushing to add layer upon layer of digital systems, automation and reporting infrastructure in the pursuit of efficiency and optimisation (the CEO of Commbank was on the news about this just last night), yet often end up more fragmented, less coherent and increasingly dependent on systems they do not fully understand or control. The technology stack improves while institutional understanding and human agency slowly erode underneath it.

I think this is why ideas like provenance, operational memory and local ownership of systems are becoming more important. There is real value in systems that help people operate with more context and continuity rather than simply abstracting decisions away from them. As AI becomes more embedded into everyday workflows, the challenge is not just building systems that are powerful. It is building systems that still leave humans with agency, understandin, control and compassion (We are More Than Machines after all).

What I liked about the encyclical is that it stays grounded in the present. It does not disappear into science fiction or spend its time speculating about superintelligence. Instead, it focuses on the systems already forming around us and asks what kind of society those systems are likely to create. One line that stood out to me was: "A society that delegates judgement entirely to machines risks forgetting that wisdom is born from conscience, responsibility and encounter." Another passage warns that "technology becomes dangerous when efficiency is treated not as a means to human flourishing, but as its replacement." Those ideas seem especially relevant right now because so much of the AI conversation still assumes that faster, cheaper and more automated automatically means better.

Whether you are religious or not, I think it is worth reading. I thought it is a surprisingly practical and well considered perspective on where AI is heading from a Pope who seems to genuinely understand the implications of where technology is taking us.