On May 5, a conference titled "From Source to Classroom: Theological Uses of Magisterium AI" was held at the Pontifical Gregorian University. Fr. David Nazar, SJ, director of Administration at the Pontifical Gregorian University, opened the event, stating that the conference aimed to embrace AI as a tool for evangelization and communication, urging the church to be at the forefront of new technologies.
The keynote speaker, Fr. Kenny Ang, research professor at the Pontifical University of Santa Croce, described Magisterium AI as a system similar to ChatGPT but specifically "fed with Catholic documents." He emphasized that one does not need to be an AI expert to engage with its development. Fr. Ang compared it to the Index Thomisticus, noting that while such tools help find specific terms, AI could find concept discussions even without exact keywords. He outlined challenges for Magisterium AI, including filtering documents, adding diverse sources beyond official Church texts, and creating a ranking of sources based on how recent and authoritative they are. He concluded that AI development requires input from scripture, theology, philosophy, and an understanding of language changes, and suggested using AI to create imaginary dialogues with saints, understand faith through art, and in pedagogy to foster deep learning and critical thinking.
The subsequent panel discussion addressed why AI should be a priority for the church. Matthew Harvey Sanders, CEO of Longbeard and creator of Magisterium AI, described AI as a disruptive technology, arguing that the church's guidance will be crucial in helping society adapt. As a compound AI system, Magisterium AI uses standard AI models but adds a special database of church documents and safety features to ensure faithful answers.
Jen Copestake, correspondent for China Global Television Network (CGTN) Europe, shared her experiences with humanoid robotics and their creators' goal of making robots indistinguishable from humans, even in terms of understanding emotions. This led her to question whether machines could have a spiritual life, underscoring the need for the church's deep involvement in these evolving discussions.
Fr. Jean Gove, research affiliate at the AI & Humanity Lab, The University of Hong Kong, noted that AI is more than just a tool and might create new forms of social inequality. He called for the church's preferential option for the poor to extend to these emerging areas of marginalization. Fr. Michael Baggot, LC, associate professor at the Pontifical Athenaeum Regina Apostolorum, highlighted the church's unique position to address both risks and opportunities of AI, suggesting that if AI creates abundance, the church can help people find deeper meaning.
On using AI in classrooms, Fr. Michael Baggot discussed changing teaching methods toward more interactive approaches like oral exams and group debates. Fr. Jean Gove talked about efforts in Malta to train teachers to use AI mainly for building skills and character, using AI to reduce paperwork and free up more time for face-to-face student contact.
Fr. David Nazar shared that AI research at the Pontifical Oriental Institute led to better human communication, with professors asking for more face-to-face meetings with doctoral students. Fr. Kenny Ang repeated his belief that AI, when used thoughtfully, can support learning without reducing person-to-person contact.
During the Q&A session, Matthew Harvey Sanders addressed concerns about potential Western bias in Magisterium AI's training data. He mentioned the efforts to include diverse theological sources and the creation of the Alexandria Digitization Hub to digitize pontifical university libraries. On the issue of AI "hallucinations," Sanders explained that Magisterium AI was developed to reduce this risk by basing answers on source documents.
Questions from the audience included the risk of diminishing the valuable process of discovery in learning and the potential for a "technocracy of knowledge." Fr. Jean Gove emphasized that final judgment must always stay with humans, stressing the importance of respecting human dignity and moral choice. Regarding academic writing in the AI age, Fr. Michael Baggot hoped AI might reduce pressure to "publish or perish," allowing scholars to focus on quality over quantity.
The distinction between a specialized "Catholic AI" like Magisterium AI and general LLMs, particularly concerning the use of synthetic data, was also a key point of discussion. Sanders stated that Magisterium AI is committed to accuracy based on actual church documents, clarifying that his team does not and will not generate artificial Catholic data, as this would damage trust. Fr. Jean Gove agreed with this position, questioning how artificially generated data could reliably reflect or lead to truth.
As the dialogue came to a close, the event concluded by underscoring the ongoing efforts and critical considerations involved in integrating AI responsibly and ethically within theological scholarship and Catholic education.
