A former papal adviser on artificial intelligence ethics, Franciscan friar Fr Paolo Benanti, has cautioned that AI-generated material risks becoming a form of ‘ultra-processed’ knowledge—marketable but lacking genuine educational value, reports Melbourne Catholic.
Fr Benanti was speaking at a panel discussion, ‘Knowledge and Humanity: Shaping AI for the Common Good’, held at Australian Catholic University in Melbourne and co-sponsored by the Italian Embassy. The professor of moral theology and AI adviser to the late Pope Francis was joined by La Trobe University’s chief AI officer Phil Laufenberg, Prof Phil Parker of ACU’s Human and Digital Futures initiative, and ACU’s head of discipline for IT, Prof Niusha Shafiabady.
Fr Paolo Benanti TOR speaking at a panel discussion, ‘Knowledge and Humanity: Shaping AI for the Common Good’, held at Australian Catholic University in Melbourne. PHOTO: Melbourne Catholic.
Opening the event, ACU Vice-Chancellor Zlatko Skrbis warned that artificial intelligence cannot be left to develop on its own terms, quoting Pope Leo XIV’s warning that technology is never neutral because it takes on the characteristics of those who devise, finance and regulate it.
“This places responsibility with us—citizens, leaders, experts, educators and scholars—to shape its development in ways that respect human dignity and serve the common good,” Prof Skrbis said. He cited the Pope’s challenge to universities to ensure “the integration of knowledge, cultivating both the capacity to connect and synthesise knowledge in order to grasp complexity, and the skills necessary to verify facts”.
During the panel discussion, Fr Benanti addressed OpenAI’s recent release of hundreds of previously unsolved mathematical proofs, questioning the value of choosing something humans do not understand to effectively show off the capacity of its model.
“A lot of scientists that try to analyse it say that it’s a little bit messy,” he said. “It looks to me much more like the Rosetta Stone than a step forward in science, in the sense that … there is something inside, but for now no one has decoded what [it] is.”
He compared the output to ultra-processed food: “We take something that is natural … We split it, and we have these super-fatty, salty sugar chips that are really super-palatable, but not at all nutritious. If this is the first step in an ‘ultra-processed education’, what will [happen]? We have a superfluency, but no education at all.” Universities exist to handle complexity, he said, and must resist oversimplified answers.
Prof Niusha Shafiabady continued the theme of learning and understanding, saying AI could not come up with anything that truly advances knowledge, giving as an example Einstein’s theory of relativity.
“Creation is something based on some kind of consciousness, and AI is not conscious,” she said. “When something is pretend, simulation, or pretending to understand, it’s different from being able to create things, being able to innovate honestly, creating the framework from the very beginning.”
The overarching message from the panel was that AI can be (and is being) used for the common good. That potential is the focus of ACU’s approach to AI, Prof Shafiabady said. “We teach the students to use AI ethically and responsibly.”
Fr Benanti described his own classroom use of AI at the Pontifical Gregorian University. He generates podcasts, infographics and short cinematic videos to extend teaching rather than replace it.
“The question was: May I use AI not to replace but to extend that time [with students]?” he said. “So AI can make my lecture more interesting. It’s a different form of teaching. It’s enhancing the teacher … augmenting the surface of contact between me and them.”
Mr Laufenberg said La Trobe’s approach is that “AI needs to happen with us and not to us”. The university works backwards from student, research and community goals rather than adopting tools for their own sake. It is redesigning curricula so graduates become ‘AI capable’—able to use the technology, aware of when not to use it, and retaining critical thinking and judgment.
Prof Parker said ACU is redesigning computer science programs to embed digital fluency and prepare graduates from all disciplines for tech and non-tech workplaces. “Universities exist not just to produce graduates to work in the labour market but—to steal a phrase from philosopher Martha Nussbaum—we’re there to ‘cultivate humanity’. And so, the critical aspect is going to be: How can we ensure … that the curriculum is continuing to push forward to cultivate humanity, to create good citizens?”
On the wider question of the common good, Fr Benanti argued that many AI systems embed hidden political choices through labelled training data.
“This base truth now is hidden in the building process of the AI,” he said.
As an example of an undeclared bias with real consequences, he cited pulse oximeters, which measure blood oxygen levels and heart rate, and which produce higher error rates on non-white skin. Why, he asked, can we embed such a bias inside this kind of complex machine “without any duty of transparency? This is a core matter of rule of law”.
During audience questions, Angela Nolan, group CEO of Mercy Health, asked how organisations were holding themselves accountable, citing the Rome Call for AI Ethics.
Fr Benanti said the Rome Call sets minimum principles rather than universal values.
“From the blueprint to the building, or from the scroll to the music, you need to pick the instrument,” he said, adding that success would come when organisations no longer needed the framework because they had their own policies in place.
Responding to a claim that AI had reduced the reality of human rights, he rejected a simplistic yes-or-no response to the question of whether we should use AI.
“I find the public debate on AI a little bit delusional,” he said, arguing instead for rules that manage risk, as with other potentially harmful technologies. AI is subject to law, “not above it”, he said.
The better path, he concluded, is to ground the discussion in shared reason rather than what divides people. “We have to find what unites us.”
This article was published in Melbourne Catholic, the publication of the Archdiocese of Melbourne.