Anthropic Co-founder Outlines Ethical Challenges of AI at Magnifica Humanitas Event
During the recent Magnifica Humanitas event, an Anthropic co-founder addressed the pressing ethical concerns surrounding the rapid development of artificial intelligence. The presentation focused on three primary challenges, including the presence of bias within training data and the difficulty of establishing accountability for automated decisions. Furthermore, the speaker emphasized the significant societal impact that widespread AI deployment could have on the human experience. The discussion underscored the necessity for responsible development practices to ensure that technology remains aligned with human values. This event highlighted the ongoing dialogue between technological advancement and the inherent fragility of the human person.
4 months ago
The Vatican’s presentation of Pope Leo XIV’s first encyclical Magnifica Humanitas brought together theological reflections on the human person and concrete concerns about artificial intelligence. Anthropic co‑founder Christopher Olah used the occasion to highlight three major ethical challenges posed by AI, while the encyclical itself frames technology within a Catholic anthropology that balances human glory with fragility. 1 2
The encyclical was unveiled on 25 May 2026 in the Vatican’s Synod Hall before a mixed audience of scholars, diplomats and religious leaders. Pope Leo XIV emphasized that safeguarding the human person is the central task of the document, asking whether AI “helps human beings become more fully human before God and one another.” 1
The text repeatedly invokes a “covenant between glory and fragility,” portraying the human body as the locus where divine dignity and creaturely limitation coexist. Biblical images of the Tower of Babel and the rebuilding of Jerusalem are used to contrast a technocratic “Babel syndrome” with a collaborative, communion‑oriented “Nehemiah” path. 1
During his remarks, Olah outlined three areas where AI raises urgent moral questions. 2
Olah warned that AI development is concentrated in a few wealthy nations and could displace large numbers of workers. He called the lack of mechanisms for fair global distribution of AI’s economic benefits a “moral imperative of historic proportions.”
He asked what it means for individuals, families and societies to flourish in an AI‑rich world, noting parental anxieties about children’s mental health and widespread uncertainty about future employment. Olah urged the Church, with its long tradition of reflecting on human dignity, to contribute its perspective.
Olah described internal structures of advanced models that “mirror results from human neuroscience” and exhibit functional analogues of emotions such as joy or fear. Because these phenomena remain mysterious, he stressed the need for continuous moral oversight beyond commercial or geopolitical pressures.
The encyclical’s theological framework links the ethical concerns to the Catholic view of the human person:
Olah’s three challenges map onto the encyclical’s call for a human‑centered assessment of AI:
Both the Vatican presentation and Olah’s remarks converge on the principle that AI must serve the covenantal relationship between human glory and fragility, rather than replace it. 1 2
Assess AI ethics through Catholic teachings on human dignity and responsibility
Catholic social teaching and moral anthropology offer a clear way to assess AI ethics: AI must be evaluated by whether it safeguards the inalienable dignity of the human person and advances the common good, with real human responsibility guiding development and use.
Catholic teaching does not treat technology— including AI—as intrinsically evil. Rather, it insists that technology’s meaning and moral quality depend on whether it is oriented to the person and the moral order inscribed in human nature. Pope Leo XIV frames this as a moral duty in an age where “human dignity is threatened by new forms of dehumanization,” requiring us to “remain profoundly human” and to let progress serve communion, not separation.
A Catholic assessment begins with anthropology and social ethics: respect for the human person entails respect for “fundamental rights” that flow from dignity intrinsic to the person.
From that follows a decisive political-moral claim: these rights are prior to society and form the basis of the moral legitimacy of authority; a society that refuses to recognize them “undermines its own moral legitimacy,” reducing authority to force if it will not recognize rights.
Implication for AI: any AI system that predictably erodes fundamental rights—through surveillance without moral limits, discriminatory profiling, or manipulation of persons—cannot be justified merely by efficiency or technical success, because the dignity-based order comes first.
The Catechism connects dignity to social obligation: the dignity of the human person requires pursuing the common good, and people must create/support institutions that improve conditions of human life.
It also grounds participation in dignity: “Participation” in promoting the common good is inherent to human dignity, with each person participating “each according to his position and role.”
Implication for AI: ethical evaluation must include whether AI expands or restricts genuine participation in social life (work, education, public communication), and whether it strengthens institutions toward the common good rather than privatized gain.
Pope Leo XIV explicitly links AI ethics to moral discernment: the guiding criteria are “the primacy of the human person,” so that “it will always be human intelligence, with its conscience and freedom, that guides technical innovations and responsibly determines their use and limits.”
This means that AI cannot be treated as morally self-authorizing (“the model decided”), because Catholic ethics makes human moral responsibility foundational—especially when power concentrates in digital systems.
Pope Francis proposes that we develop a “new discipline” like “the ethical development of algorithms” (algor-ethics)—aimed at ensuring a competent and shared review of processes that integrate relationships between humans and today’s technology.
He grounds this review in principles of Catholic social teaching—dignity of the person, justice, subsidiarity, and solidarity—so that technology is “at the service of every individual… and of all people, without discrimination or exclusion.”
Implication for AI: ethical responsibility is not only individual; it is institutional and procedural. Catholic ethics calls for shared review and competent governance of algorithmic design, deployment, and use.
Pope Leo XIV adds a direct responsibility claim: developers have “a particular ethical and spiritual responsibility, for every design choice reflects a vision of humanity.” Developers are called to embed values with “transparency, responsibility toward affected communities,” and “careful attention” to ensuring what is cultivated is a “genuine good.”
Implication for AI ethics: building AI is not merely engineering; it is moral authorship. Transparency, impact assessment, and accountability to affected persons are not optional “extras.”
Pope Leo XIV identifies the central danger: “new forms of dehumanization.”
In Catholic terms, this is not reducible to “bad UX” or “bias metrics.” It is a failure to keep the person central—turning human beings into objects for prediction, control, or manipulation.
Pope Francis warns that we cannot accept a “digital world designed to exploit our weaknesses and bring out the worst in people.”
He also insists that communication advances should guide people toward generous encounter, honest truth-seeking, service, closeness to the underprivileged, and the promotion of the common good.
Implication for AI ethics: systems that intensify polarization, emotional manipulation, or dependency—especially where they are designed to exploit vulnerabilities—fail a dignity-based moral test even if they technically “work.”
In Fratelli Tutti, Pope Francis describes a broader diagnosis: alongside scientific and technological advances there can exist “great and valued as they are” paired with “a moral deterioration… and a weakening of spiritual values and responsibility,” contributing to isolation and desperation.
Implication for AI ethics: if AI accelerates power without a shared moral roadmap, it can intensify fragmentation rather than communion—contradicting the Church’s insistence on the common good and participation.
Pope Leo XIV provides an explicit set of judgment/deliberation criteria in the context of digital power and algorithms: they must assess whether digital infrastructures and algorithms (1) foster participation and responsibility, (2) protect the vulnerable, (3) ensure fair access to opportunities, and (4) remain directed toward the good of all.
These criteria can be translated into an evaluation framework:
A dignity-based test includes whether AI protects those most at risk of harm and exclusion, and whether it ensures fair access to opportunities.
Ethical AI should strengthen participation and responsibility rather than replace them. Catholic participation is voluntary and generous, rooted in dignity.
AI must remain directed toward the common good and not become an engine for injustice or arbitrary power.
Even when AI expands capabilities, conscience and freedom must guide innovations and determine legitimate use and limits.
While not limited to AI, the Church’s moral reasoning about technological intervention provides a transferable principle: technology offers many possible interventions, but “careful moral discernment is needed” to decide what should be realized to promote the good of the human person.
A crucial tenet is that there is an “order in the natural world” and in human nature that must be respected, because human nature is a gift from the Creator, not something people own or can manipulate without limit.
The same source insists: neither patients nor physicians nor researchers have “unlimited rights over the body”; the patient is “the user and not the owner.”
Implication for AI: when AI is used in areas touching human embodiment (healthcare decisions, identity/biometric systems, reproductive or body-related technologies), ethical assessment must respect the human person’s created order and its limits—not treating the person as a modifiable object.
Catholic AI ethics evaluates AI according to human dignity (rights prior to society), responsibility (human conscience and freedom guiding limits), and the common good (participation, protection of the vulnerable, fair access, direction toward the good of all).