The Vatican’s response to artificial intelligence began a decade ago in the private library of a religious order in central Rome, where a group of top Silicon Valley executives gathered with senior Catholic officials below shelves lined with antique books spanning centuries of Christian thought.
The issues raised culminated this week with the publication of a landmark encyclical by Pope Leo XIV that summoned the world’s 1.4 billion Catholics to work towards the ethical governance of a technology with a disruptive potential akin to the industrial revolution.
The meeting grew out of an informal encounter at a Bay Area conference in 2016, when a group of Silicon Valley executives including the LinkedIn co-founder and early Open AI investor Reid Hoffman approached a French priest, Éric Salobir, and asked how they could contact the Vatican.
“They wanted to alert us to something that was coming,” remembers Bishop Paul Tighe, an Irish senior Vatican official who was present at that first gathering in Rome.
“We had some very senior people, and it became very clear that this was genuinely taking off, and they were surprised at the rate of development,” Tighe remembers.
“That’s the thing that struck me at the time – the pace at which it was developing, and also the range of areas where it would be relevant . . . You realise that this wasn’t just going to have an impact in the narrow area of AI, but was going to be transformative.”
The tech executives were interested in involving faith traditions in general, not the Catholic Church exclusively. But conveniently, Catholicism had a “corporate headquarters” that they could approach.
That meeting in the Dominican library was the first of what would come to be known as the Minerva Dialogues, annual closed-door meetings between the Vatican and Silicon Valley named after the adjoining church of Santa Maria Sopra Minerva, built in the 13th century on the ruins of a Roman temple.
It’s a site that is heavy with history of the church’s relationship with science and technological development. Somewhere within the complex in 1633, church inquisitors forced the father of modern astronomy, Galileo Galilei, to renounce his belief that the sun rather than the Earth was at the centre of the universe, following his trial for heresy.
Those attending the Minerva Dialogues quickly came to see AI as a technological development that offered huge promise to humanity, but also something that carried profound risks.
“I remember being at a session where they were showing the potential benefits of AI in terms of diagnostics, in terms of individualised treatment plans,” Tighe says.
“But then it struck me at the time: yes, but if we don’t fix the equality of our healthcare systems, this is going to be great for some, and really just make more pronounced the inequalities.
“We gradually came to the realisation that AI has this extraordinary ability to magnify both what is best about humanity, and also our far less good tendencies.”
Key interlocutor
Held each year since that first meeting in 2016, the discussions brought together figures such as former Google chief executive Eric Schmidt and chief technology officer at Microsoft Kevin Scott with senior Vatican officials, Catholic theologians and philosophers.
Tighe – mostly due to his native English, he says – ended up becoming a key interlocutor and co-ordinator between the Vatican and Silicon Valley in the process.
It was a strange, unexpected kind of return for the Meath-born prelate, who grew up partly in Sligo, where his father worked for IDA Ireland at a time when the agency began looking towards Silicon Valley as a potential source of investment into Ireland.
“It was funny, some of the names, even the name of Silicon Valley itself and some of the companies, I’ve been familiar with from home,” he says.
Tighe is a former teacher who worked in communications under then archbishop of Dublin Diarmuid Martin. He was called to the Vatican in 2007 at a time when the Holy See was digitalising its public outreach and was involved in setting up the papal Twitter account.
Since 2022, he has been the secretary – second in command – of the Dicastery for Culture and Education, the Vatican equivalent of a ministry.
Over the years, his incongruous presence has been noted by media reporting from tech conferences such as the Web Summit and South by Southwest.
His work involved frequent travel between the Vatican and Silicon Valley, where a parallel process to the Minerva Dialogues was taking place at the Markkula Center for Applied Ethics, part of a Jesuit university in Santa Clara, California, and home to a research group on the ethics of AI.
“I met Fr Brendan McGuire, who was a parish priest in that area, an Irish priest who had worked in technology,” recalls Tighe. McGuire “had many parishioners who were working in the Silicon Valley companies, who were anxious to harmonise their work and their faith”, he says.
The issue soon came to the attention of the man at the top.
“Pope Francis, I remember he called me in for a meeting about 2018,” Tighe remembers. “He said he’d had a visit from a number of business people who had told him that the church needed to begin to have a more kind of deliberate reflection on AI. He wasn’t a technologist, you know, but he said, ‘Look, try and develop these contacts and keep in touch’. And then, gradually, he began to speak about it.”
The turning point came with the launch of the AI text generator ChatGPT in 2022, when the topic of artificial intelligence became a primary focus of the world’s attention.
“I greatly value this ongoing dialogue,” Pope Francis told participants of the Minerva Dialogues in a meeting in 2023, expressing hopes for “a serious and inclusive global discussion on the responsible use of these technologies”.
In an address to a G7 meeting in Italy a year later, the pope described AI as having the potential to make great advances for humanity but also to worsen injustice, describing the algorithms behind the systems as “neither objective nor neutral”.
A few months later, Tighe’s Dicastery for Culture and Education published a joint document with the Vatican’s doctrinal department, the Dicastery for the Doctrine of the Faith.
Antiqua et Nova laid out, in 117 clauses, the culmination of the Vatican’s reflection on AI, its implications, and the ethical concerns it posed in various areas: employment, healthcare, education, misinformation, privacy, surveillance, the environment and warfare.
It stressed that there was no equivalence between human intelligence and its “imitation”, artificial intelligence. “No AI application can genuinely experience empathy,” it read. “AI’s advanced features give it sophisticated abilities to perform tasks, but not the ability to think.”
Within the first hours of the election of the new pope just over a year ago, it was already evident to close observers that artificial intelligence would be a key concern to his papacy. The clue was his choice of name: Leo.
The last pope named Leo was a giant in the modern history of the Catholic Church, remembered for shepherding it through a time of epochal change and for authoring what would come to be known as the church’s “social doctrine”.
The crucial text was Leo XIII’s 1891 encyclical Rerum novarum, which addressed the plight of the class of urban poor created by the industrial revolution.
It defended the right of workers to form trade unions, but also spoke in support of private property, a middle way between unfettered capitalism and then-rising radical socialism that became influential far beyond Catholicism.
The current Pope Leo symbolically signed his first encyclical on the 135th anniversary of Rerum novarum, and began it by saying he wished to “add my own voice” to the tradition it began.
Its launch this week was highly unusual, featuring addresses by a panel of speakers including the co-founder of AI company Anthropic, Christopher Olah. It was a first for the pope himself to make an address, lending the work additional prominence.
Epochal change
The encyclical describes AI as a development that is still evolving – “any statement regarding AI risks becoming quickly outdated”, it notes – but that clearly marks an “epochal change”.
It flags the risk of the “social calamity” of mass unemployment, worsened economic inequality if the technologies are kept in the hands of the few, the potential for discrimination in algorithms, and AI weaponry that could lower “the moral threshold of conflict”.
But the encyclical urges readers not to become fatalistically resigned, but to realise their own potential. “No one is without responsibility. We all have our own areas for action,” it reads.
A primary aim appears to be to boost momentum for some kind of binding international agreement – probably through the United Nations, which is name-checked in the document – and legislation to ensure the ethical governance of AI and to ensure that the technology is harnessed for the good of humanity.
The serious challenges in the way of such an agreement are also flagged in the document, which has a section called “the crisis of multilateralism” and states that the era of postwar co-operation has given way to “a disorderly and conflict-ridden multipolarism with a prevailing sense of mistrust”.
For Tighe, the church can point the way, but others have to walk the path.
“The church can say, ‘Look, this is what humanity should be aspiring to’. We can’t on our own bring that about,” says Tighe. “It is an encouragement for people who are working for that to feel empowered.”