AI’s moral impact is best judged, in Catholic terms, by asking a single question: does this technology help the human person flourish as a person—capable of truth, freedom, conscience, love, and relationship—or does it reduce the person to a “thing,” a “resource,” or a manipulable object? That human-centered criterion runs through the Church’s recent teaching on AI and clarifies both the promises and the dangers. 1 2 3 4 5

Catholic human dignity: the moral “lens” for evaluating AI

Catholic anthropology insists that the dignity of the human person is intrinsic and not earned. In Magnifica Humanitas, Pope Leo XIV distinguishes dignity levels but emphasizes that at the deepest level it is ontological—rooted in the fact that each person is willed, created, and loved by God, and therefore cannot be erased by failure, humiliation, or exclusion. 6

This dignity is not merely a social ideal; it grounds concrete moral obligations:

• Persons are moral subjects: “Freedom makes man a moral subject,” so when actions are deliberate, the person is truly responsible. 7
• Technologies must be subordinated to the person: the Church repeatedly warns against a “technocratic paradigm,” where efficiency, control, and profit dictate what matters and what can be discarded—reducing people to “cogs” and even creation to exploitation. 2
• AI is not human intelligence: the Church clarifies that so-called AI “imitate[s] certain functions of human intelligence,” but does not possess experience, embodiment, conscience, or moral responsibility. 2

  “They may imitate language… or even simulate empathy… but they do not understand what they produce… for they lack the…” 2

  Translatio automatica

So, the moral lens is clear: AI may be powerful, but it must remain human-centered and ethically governed so that it never displaces conscience, freedom, and responsibility as the true center of moral decision-making. 2 8

The technocratic paradigm: when AI’s structure widens inequality or instrumentalizes persons

One of the strongest Catholic critiques is not that AI is inherently evil, but that it can accelerate a deeper cultural shift: technology becomes the standard for judging everything. Pope Leo XIV explains that this technocratic paradigm risks turning:

• human beings into mere instruments for results,
• participation into something shaped by opaque corporate control,
• and vulnerability into something treated as discardable. 2 3

In his encyclical, Leo XIV names a specific political-moral danger in the digital context: control over platforms, infrastructure, data, and computing power may not rest with States, but with major economic actors who set the conditions for visibility and participation—often with limited public oversight. 2

This threatens human dignity in at least three ways:

• Concentration of power undermines transparency and accountability, and therefore undermines real respect for persons in society. 2
• Efficiency-first value judgments risk shifting evaluation away from the person and toward outcomes—“having more” without “being more.” 2
• Vulnerable people become easier to exclude when “participation” is filtered by systems that are opaque or optimized for profit. 2 3

Pope Francis frames this moral evaluation in terms of the aims and uses of AI: it depends not only on technical design, but also on “the aims and interests of its owners and developers” and the situations where it is employed. 8
He also warns against “throwaway culture” tendencies that can displace a “culture of encounter.” 4

Catholic conclusion: AI’s moral impact is not just “what it can do,” but what it tends to reward socially—and whether it fosters a framework of justice and fraternity or a framework of discard and domination. 2 8

Conscience and responsibility: AI must not replace human moral agency

A second major Catholic criterion concerns the boundary between tool and moral agent. Pope Francis insists that AI is “above all else a tool,” and therefore its moral effects follow from its use. 4

But he also highlights a practical moral failure mode: the temptation to draw general or anthropological conclusions from specific outputs, especially when a machine begins to influence decisions about persons’ futures. 4

He gives a concrete example involving judge-assist systems that predict a prisoner’s likelihood of reoffending. Such systems can draw on categories tied to private life and may effectively delegate “the last word” to a machine—implicitly incorporating prejudices in the data categories used. 4

From a dignity lens, the core problem is not merely technical error; it is the moral structure of decision-making:

• Human dignity requires that decisions about persons remain human—because persons are moral subjects capable of conscience and responsibility. 7 7
• Pope Francis warns against taking away peoples’ ability to decide about themselves by making them depend on machines. 7
• Pope Leo XIV emphasizes that moral discernment must ensure it will always be human intelligence, with its conscience and freedom, that guides technical innovations and responsibly determines their use and limits. 2

Pope Leo XIV develops this further through governance concerns: he argues that because of AI’s complexity and pace, there is an urgent need for both scientific deepening and “moral and spiritual discernment.” 2
And in the context of war, he states the moral non-transferability of lethal decision-making:

This teaches a general principle: where moral responsibility is involved, the chain of responsibility must be identifiable and verifiable, and those who design, train, authorize, and employ technology must be held accountable. 9

Catholic conclusion: the moral impact of AI is seriously harmful when it undermines the human person’s role as the responsible subject of moral judgment—especially by making decisions opaque or effectively outsourced. 9 4 2

Justice, rights, and peace: dignity is tested by bias, privacy, inclusion—and by limits on violence

The third criterion is social and relational: dignity protects everyone’s rights and demands justice for the whole human family, not only those favored by market power or data advantage.

Algorithmic bias and the risk of discrimination

Pope Francis notes that predictive systems can incorporate prejudices embedded in data categories; classification by ethnicity or minor past offenses can affect decisions even though human beings are “always developing.” 4
The DDF note also addresses “bias” as algorithmic or learning bias—systematic and consistent errors that can disproportionately prejudice certain groups. 7

Dignity-based evaluation therefore asks: Does AI systematically disadvantage people in ways they did not choose and cannot contest? If so, the technology fails the dignity test. 4 7

Privacy and participation

The Church’s teaching on communication and human rights insists that social systems must respect human dignity, including the right to privacy and not being exploited in the intimacy of one’s family. 10
And Pope Francis describes how AI systems may interact as chatbots and hold conversations tailored to users’ psychological and physical needs—raising a dignity question about manipulation and the shaping of dependence. 4

Inclusion, transparency, and the common good

Pope Francis repeatedly lists fundamental values that must guide responsible AI development and use: inclusion, transparency, security, equity, privacy, and reliability—and he warns that technological progress that worsens inequalities can never be “true progress.” 8
He also calls for formation in responsibility and for bodies capable of examining ethical issues and protecting rights of those affected. 8

Peace and warfare: AI cannot lower the moral threshold of violence

For dignity, peace is not secondary; it is a direct moral test. Pope Leo XIV states that developments in AI-enabled warfare make war more “feasible” and less under human control—violating the principle that armed force should be used only as a last resort in legitimate self-defense. 9
He argues that AI can also make conflict more impersonal and render it more quickly, lowering the threshold for resorting to violence and “accustom[ing] us” to treating violence as something to optimize. 9

Catholic conclusion: dignity is honored when AI strengthens inclusion and protects rights in ways that advance peace; dignity is violated when AI amplifies discrimination, erodes privacy and accountability, or accelerates impersonal, automated violence. 8 9

What safeguarding human dignity requires from AI (a dignity-based evaluation checklist)

From the Church’s teaching, a Catholic evaluation of AI’s moral impact should consistently include these questions:

• Human-centered governance: Is human conscience and freedom truly guiding the system, including its limits and purposes? 2
• Accountability: Can the chain of responsibility be traced—designers, trainers, authorizers, and users held accountable—especially for irreversible harms? 9
• Transparency and participation: Are affected persons able to understand, contest, and participate rather than being excluded by opaque systems? 10 2
• Equity and anti-bias: Does AI avoid systematic prejudice and respect the fact that persons are more than past data? 4 7
• Privacy and protection of the vulnerable: Does AI protect human interiority and avoid exploitation of private life? 8 10
• Peace-oriented limits: Are AI uses constrained so they cannot lower the moral threshold for violence or remove human control over lethal decisions? 9 9
• Anti-technocratic discernment: Does AI serve the common good rather than dictating what can be discarded under efficiency/profit logic? 2 2

Conclusion

Catholic teaching evaluates AI’s moral impact by whether it safeguards the equal and inalienable dignity of every human person and preserves the person’s unique role as a responsible moral subject. When AI strengthens conscience, justice, inclusion, transparency, privacy, and peace, it can serve the common good; when it accelerates technocratic domination, outsources moral judgment, entrenches bias, or lowers barriers to violence, it risks turning persons into instruments rather than recognizing them as persons. citeturn0file7turn0