Guest Post: How does Decision Intelligence help to answer Pope Leo’s AI challenge? (Part 2 of 2)

In Part 1 of this two-part series, we explored how Decision Intelligence (DI) can support prudence by helping us reason explicitly about how actions causally lead to outcomes. Here, we examine how that practical discipline can help advance Pope Leo’s call for technology to serve humanity.

The connection begins with a challenge in Magnifica Humanitas: “It is not enough to invoke ethics in the abstract.” In paragraph 106, Leo calls for prudence, rigorous evaluation, and institutions capable of governing AI’s effects. Encyclical, §106

What would it take to carry those commitments into practice? In using technology, an organization or individual needs to explain what it hopes to achieve, how its choices could produce those outcomes, whose lives could be affected, and who will take responsibility.

Decision Intelligence provides methods for making stakeholders, outcomes, and consequences explicit. It connects human values to practical choices through causal reasoning that people can examine, challenge, and revise. Here’s our interpretation of how DI can help put Leo’s concerns into practice.

1. Human values must shape actual choices.

In paragraph 104, Leo explains that technical systems embody priorities through what they measure, ignore, and optimize. Ethical scrutiny must therefore reach into their design. Encyclical, §104

Consider a public agency deciding how to introduce AI into its benefits application process. It hopes to reduce costs and process applications faster. But a commitment to human dignity raises further questions: Will eligible people receive support sooner? Could the process exclude people who cannot navigate it? Will applicants be able to obtain help or challenge an error?

DI helps translate these concerns into explicit goals and constraints. Timely access to support becomes an outcome to pursue. Preserving a meaningful route to assistance or appeal becomes a requirement against which proposed changes can be assessed.

This also changes the scope of the decision. A system that saves the agency money might impose additional costs on applicants, families, or community organizations. Those consequences belong in the reasoning, even when they fall outside the agency’s budget.

Human values then influence which alternatives are considered and how they are evaluated. DI helps make human dignity and the common good consequential to the choice itself, giving people a way to challenge proposals whose apparent success depends on overlooking harm.

2. Choices require explicit reasoning about consequences.

Leo’s call for prudence and rigorous evaluation in paragraph 106 raises a practical question: how can we assess whether a proposed action will produce the good we intend? Encyclical, §106

For our fictional agency, “introduce AI” is too broad a proposal to evaluate well. The consequences depend on specific choices: which tasks to automate, whether to retain in-person assistance, how to handle unusual circumstances, and when staff should intervene.

DI makes the causal reasoning behind those choices explicit. Removing in-person assistance could make some applications harder to complete. Incomplete applications could delay access to benefits. Those delays could worsen food or housing insecurity.

A different choice might produce a different chain of effects. Using AI to reduce routine administrative work could free staff to assist people with complex needs, improving access while also reducing delays.

Neither chain should be accepted simply because it sounds plausible. Each contains assumptions to examine using evidence, models, and the knowledge of people familiar with the process. A causal decision diagram makes these connections visible so that people can identify gaps, compare alternatives, and investigate uncertainty.

DI gives practical support to prudence by helping people assess which actions are likely to advance their values – and uncover how well-intended choices could undermine them.

3. Affected people must help shape decisions.

In paragraphs 107–109, Leo connects AI governance to public deliberation, participation, and subsidiarity. Communities must have a meaningful ability to influence choices and corrections. Encyclical, §§107–109

Link stresses that decisions in complex systems “…cannot be limited to highly-trained experts. They must be democratized and ubiquitous, inviting citizen contribution.” The DI Handbook advises, “Make sure to check with all potential stakeholders to ensure the levers, outcomes, and externals they consider most important are reflected.”

The practical connection is that participation can improve – and change – the reasoning on which a decision rests.

Applicants might explain why a required document is difficult to obtain. Frontline staff might identify circumstances that standard forms fail to accommodate. Community organizations might reveal consequences that the agency’s performance data do not capture.

DI provides a shared structure in which this knowledge can influence the understanding of how actions lead to outcomes. Participants can see whether their concerns are represented, question assumptions, and propose alternatives. Disagreements become visible: people may value different outcomes, or disagree about what a proposed action will cause. Those differences require different conversations.

Subsidiarity also raises the question of authority. A local office that identifies a problem needs an appropriate ability to respond. A shared decision model can help clarify which choices can be made locally, which require coordination, and how each affects wider goals. The organization must then assign the authority and resources to act.

DI helps equip affected people to shape decisions with their own knowledge and reasoning, supporting the meaningful participation and local agency that Leo calls for.

4. People must remain accountable and able to correct course.

Paragraph 105 connects responsibility to the ability to justify decisions, monitor them, challenge them, and remedy harm. Encyclical, §105

Decision Intelligence is an enabler for transparent human decision making. Link describes connecting human decision makers to technologies and disciplines that turn invisible decisions into visible ones.

That visibility gives accountability something concrete to examine. A decision record can show what was chosen, which outcomes mattered, what consequences were anticipated, and why decision makers believed the action would work.

For the benefits agency, accountability also requires monitoring beyond the headline measure of processing speed. Are more people abandoning applications? Are some groups experiencing longer delays? Is the expected improvement in access actually occurring?

Comparing actual outcomes with the original reasoning helps identify where an assumption failed or a consequence was missed. It also provides a basis for changing the process and assigning responsibility for corrective action.

By making choices, assumptions, and consequences open to scrutiny, DI helps people demand explanations and corrections – and helps responsible organizations learn before harm grows.

DI developed from the practical challenge of helping organizations make better decisions in complex environments. The encyclical is rooted in Catholic social teaching, human dignity, and the moral obligation to serve the common good. Their connection becomes useful when those moral commitments guide the reasoning behind specific actions.

DI can contribute to this work alongside the institutions, oversight, and human responsibility Leo calls for. It helps people make values explicit, investigate consequences, include affected communities, and assess whether their choices are achieving what they intended.

A commitment to human flourishing becomes actionable when we can explain how our choices are expected to advance it – and take responsibility for changing them when they fall short.