In the work we do to support healthcare organizations, we see daily how AI tools are already transforming the industry. Pope Leo XIV’s first encyclical, Magnifica Humanitas, recognizes the commonsense proposition that such transformations can have both positive and negative aspects.
While this powerful technology can “often surpass human intelligence in speed and computational capacity, offering tangible benefits across many fields," the benefits may not be shared equally or equitably: “To think that new technologies will automatically benefit everyone is to ignore the evidence.” And so Pope Leo XIV challenges all of us who adopt, use, develop, and deploy AI technology to stop and think beyond “what” we want these tools to accomplish and to also consider “how” they will do so. As illustrated by numerous examples in the 42,000-word encyclical, the second question is critical to preserving our humanity and shaping the world for future generations.
Granted, Magnifica Humanitas does not carry regulatory force or impose any strict obligations on healthcare organizations. Yet that doesn’t mean it has no impact on—or lessons for—the industry at large, particularly as AI adoption continues to surge.
Broadly speaking, the papal letter encourages a deliberative, governance-first approach to AI adoption and use that places human dignity and compassionate care at the center of the conversation. It’s logical to assume that Catholic health systems may be among those most inclined to consider this guidance in shaping institutional policy. We suggest, however, that their secular or other sectarian peers may do well to consider this guidance as well.
Magnifica Humanitas, or "On Safeguarding the Human Person in the Time of Artificial Intelligence," brings to light several critical AI considerations for healthcare leaders. Here are three key takeaways:
A prudently deliberative, governance-first posture to AI may win in the long run
For organizations deciding whether to deploy a new AI tool, the encyclical provides direct doctrinal grounding for a thoughtful, governance-first posture.
Paragraphs 105-106 of the encyclical say that in every AI system "responsibility must be clearly defined at every stage,” noting that “prudence, rigorous evaluation and even, at times, a slower pace in adopting AI" are part of "responsible care for the human family," and that "robust legal frameworks, independent oversight, informed users" are required.
The encyclical’s foregoing paragraphs contain a direct warning that those who design AI systems and processes inherently impose their own moral code. Those who adopt or employ such systems are therefore obligated to confirm affirmatively a proposed AI deployment’s alignment with the values of the organization. AI adopters must exercise active discernment and ongoing monitoring and governance.
The encyclical also cautions that human-centered AI governance is the standard to which to aspire, rather than approaches reducing decisions to solely machine-based utility or efficiency. As AI tools become stronger, there is ever greater temptation to accept their output without challenge, question or verification.
In healthcare, where life-and-death situations are part of daily operations, such unanticipated outcomes could have dire consequences. As such, the more durable path—incorporating standard AI governance standards—may include documented policies and procedures to ensure meaningful “human in the loop” and clarify how and why AI is being used to assist and inform the human determination.
Auditing AI tools for bias, discrimination, privacy—and ethics—is crucial
Throughout the encyclical, Pope Leo XIV nods to healthcare organizations’ obligations to audit AI tools for bias, discrimination, privacy protections and positive values.
Paragraphs 103 and 171, for instance, reference the risk that AI-mediated decisions affecting "access to credit, employment or essential services" can discriminate against vulnerable persons behind a "veneer of neutrality and objectivity."
In the context of a healthcare system, this could include deploying administrative AI tools governing patient data, risk scoring, resource allocation and operational triage. Relevant to such deployments, both healthcare law and Catholic social teaching impose anti-discrimination obligations grounded in concepts of human dignity. Magnifica Humanitas identifies health data as a particularly sensitive concentration of power, because those who control it may effectively determine how medicines, investments and protections are allocated. That concern aligns with legal duties already reflected in HIPAA, Section 1557 of the Affordable Care Act, state anti-discrimination laws, and mounting AI-specific regulatory guidance around auditability, explainability and bias testing.
Initial audits and system design should also factor in positive values and ethics. Paragraphs 102-104 insist that when AI enters people’s lives, it "touches on rights, opportunities, status and freedom," and that automated systems cannot exercise compassion, mercy or the recognition that people are capable of change. Consequently, delegating decisions to such systems may create "new forms of exclusion.” For healthcare organizations, this suggests an obligation to scrutinize not just whether a clinical AI tool is being used for good purposes, but also whether the values embedded in its design and training data are consistent with the inalienable dignity of the patient.
That extends to vendor relationships, too. Those that require algorithmic transparency, bias auditing and data stewardship commitments from third-party providers are aligning with what the encyclical demands and where many state AI regulatory frameworks are moving toward anyway.
Human beings, connection and accountability must remain at the center of care
Magnifica Humanitas warns that the artificial imitation of care becomes especially dangerous where real relationships are already thin, because the gradual erosion of the desire for genuine human connection presents a potentially deeper harm. In healthcare law, that concern maps directly onto patient rights doctrines involving communication, language access, disability accommodation and the therapeutic relationship, where regulators increasingly ask whether automation has displaced the human engagement vulnerable patients require.
AI communication tools, for example, may reduce some staffing burdens and improve consistency and scale. But if they later erode trust, increase complaints, contribute to miscommunication or trigger regulatory scrutiny, the cost of redesigning systems, retraining staff and rebuilding confidence will exceed the savings of rushed adoption.
The importance of human dignity and relationships runs through every substantive chapter of the encyclical: technology, it notes again and again, must serve and enhance human processes and relationships rather than displace or render them secondary, and human accountability should remain paramount “at every stage." Those organizations that can articulate, for every AI tool they deploy, precisely how human judgment remains primary, how accountability is traced, and how the dignity of the individual patient is protected from the logic of optimization, will have understood the encyclical's deepest challenge. Those who cannot have allowed efficiency to become, in the pope's phrase, "the ultimate measure of value."
Taking the long view on AI in healthcare
The approach outlined by Magnifica Humanitas may require a little faith (pun intended) on the part of decision makers to look beyond the short-term financial impacts and take in the longer view.
The healthcare organizations that draw on deliberative, long-term AI governance principles extolled in Magnifica Humanitas (and elsewhere, such EU AI Act — Regulation (EU) 2024/1689), NIST AI Risk Management Framework (NIST AI 100-1, 2023, as supplemented in 2026) and OECD AI Principles (OECD/LEGAL/0449; adopted 2019, revised 2024), for example), are more informed and prepared for our evolving world. They ask harder questions of vendors and demand explainability in clinical tools. They avoid the temptation of rapid adoption that can spur mistakes. Taking this path can reduce legal liability, as well as reputational and financial damages. We do not see the pope or these others as suggesting shunning or unduly delaying AI use; instead, they are calling for thoughtfully planned implementation.
Though it may require some short-term deferral, the long-term impacts of a deliberate, purposeful approach to AI may very well be worth it. It is that sort of approach that will assure that the quality of each healthcare organization “is measured not by the power of its means, but by the care it is able to offer, by its ability to recognize the other as a face not merely as a function."