The Catholic sector should lead on AI, together

“Welcome to the second Gilded Age.” So began a recent essay in Time Magazine describing our current predicament.

The first Gilded Age, which ran from 1870 to 1914, belonged to the Rockefellers, Carnegies, and Vanderbilts – men (and they were all men) of heavy industry who commanded the era-defining sectors of oil, steel and transportation, and in turn amassed fortunes and influence so vast that they alarmed the very governments who were meant to keep them in check.

This dominance exacted a heavy cost, as their vast fortunes were built on the labour of men, women, and children who worked punishing hours for low pay, in dangerous conditions, and with no security or voice.

It was in this context that Pope Leo XIII issued Rerum Novarum in 1891, which insisted on upholding the human dignity of workers as persons owed a just wage and humane conditions, and to never be treated as mere instruments for making money.

Moreover, an upheaval of that scale, he argued, could not be left to market forces to resolve. Rather, the conditions of the first Gilded Age called for employers, governments and communities to act collectively, and for people to join together rather than face the new economy alone.

Today we face a similar problem. A small number of men – Altman, Amodei, Hassabis, Musk, and Zuckerberg – have amassed a huge concentration of wealth and power through their control of artificial intelligence (AI), a tool which – alongside quantum computing – may come to define the twenty-first century. In so doing, they now command the ability to make decisions that affect millions with little accountability, oversight, or control.

We saw this in April when Anthropic – the AI lab behind Claude – delayed the release of their latest model ‘Mythos’ after it proved so adept at uncovering flaws in software that it posed a major cybersecurity risk to critical infrastructure, such as the systems that underpin our hospitals. Anthropic continue to keep Mythos under lock and key, restricting access to a few hand-picked individuals and organisations.

It is perhaps little wonder, then, that Leo XIV’s first encyclical, Magnifica Humanitas, responds to our current angst around AI by taking up the question that Leo XIII first posed, asking us “what does it mean to safeguard our humanity?” in this second Gilded Age.

Magnifica Humanitas offers several thoughtful answers. First and foremost, we cannot – and should not – conflate this form of ‘intelligence’ with our own. As Leo XIV writes, “[s]o-called artificial intelligences do not undergo experiences, do not possess a body, do not feel joy or pain, do not mature through relationships and do not know from within what love, work, friendship or responsibility mean”.

Yet these are precisely the capacities on which good care depends. To recognise the vulnerability and dignity of a patient, to be present in suffering, and to exercise moral judgement in the therapeutic relationship are all reflected in the human side of the healing vocation, and they are not replicable by any machine.

However, Leo XIV’s deeper concerns appear to lie beyond what these machines cannot do. AI is fuelling the techno-utopianism of Silicon Valley, which treats the human as “something to be perfected or surpassed”, and in turn frames our limits – illness, old age, suffering, and vulnerability – as faults to be corrected, rather than a natural part of our finite being.

Another worry is that AI tools can perpetuate existing biases within their training data, in turn potentially widening the very inequities our health system seeks to close. In skin cancer screening, for example, the imaging datasets used to train AI tools overrepresent lighter skin tones and perform worse for darker ones.

Such exclusion of the vulnerable can hide behind what Leo XIV calls a “veneer of neutrality”, which makes injustice harder to see and to challenge. These risks are particularly concerning for the care economy, where those most dependent on AI-assisted decisions are often those least able to contest them.

None of this makes AI inherently bad. Used appropriately, it can reduce the burden of paperwork, yield more accurate diagnoses, and help providers reach under-served communities. Its promise, though, depends less on what it can do than on whether we have the wisdom to direct it well, asking not only what we are optimising, but for whom, and at what cost.

Late last year, Dr Xavier Symons of the Plunkett Centre for Ethics wrote an excellent article on how the Catholic sector can lead the ethical use of AI, and why it is well positioned to do so.

I would go further, arguing that we should take the lead, together.

We should because Catholic organisations share a common mission that offers a clear alternative vision for how to best use AI in the health system – one that is grounded in the dignity of the person and the common good (rather than profit or efficiency alone), and is underpinned by principles such as subsidiarity, and the preferential option for the vulnerable.

In an AI industry that is moving at breakneck speed, this approach is not only refreshing but necessary; for the principles of prudence and discernment – so central to the Catholic tradition – are needed now more than ever.

And we should do this together because – much like in the first Gilded Age – the conditions that we face today require us to respond collectively, in a collaborative manner that amplifies our voice. In doing so, we can ensure that we set common terms with those who build these tools, that no technology enters our care until its impact on the most marginalised has been assessed, and that the question of how AI is used, and for whose benefit, is not concentrated in the hands of the few.

There is no doubt that AI will change how care is delivered across this country. Whether it is allowed to change what care is depends on us, and on whether the sector can collectively lead on the safe and responsible use of AI. We should meet this second Gilded Age as Leo XIII met the first: certain that the person comes first, and that no one should face it alone.

CHA’s position paper on the Safe and Responsible use of AI is now available on our website