The False Binary: AI, Human Dignity and the Governance Question Business Leaders Cannot Avoid

A note on what follows: This piece draws on the first encyclical of Pope Leo XIV, Magnifica Humanitas, released May 25, 2026. That reference is not an appeal to religious authority, and this is not a piece of Christian writing. It is an acknowledgment that the conversation about AI's impact on work, society, and human dignity has reached the apex of global leadership, from heads of state to international bodies to, now, the papacy. Whatever one's religious affiliation or lack thereof, the encyclical offers a set of analytical frameworks and governance principles that are grounded in broadly shared human values and are directly applicable to the decisions organizations face today. The argument here stands on its own terms. The encyclical is used because it is rigorous, because it is timely, and because it says some things about AI and human dignity that the management literature has not said as clearly.

‍A Familiar Pattern

The structure of this debate is not new. History offers a useful mirror.

In 1891, Pope Leo XIII issued Rerum Novarum, a landmark document that confronted the social upheaval of industrial capitalism. The encyclical arrived at a moment when two hardened camps dominated the discourse about industrialization. One camp treated the market and its technologies as essentially self-correcting, arguing that interference with their operation was both impractical and immoral. The other camp argued that capitalism was irredeemably destructive to human dignity and must be replaced wholesale.

Leo XIII refused both positions. He acknowledged that industrial technology was real, consequential, and not going away. But he insisted that neither the technology nor the economic arrangements it produced were morally neutral. Workers had rights that predated market outcomes. Those rights included just wages, humane conditions, and the ability to organize. The state had a legitimate role in protecting those rights when market forces failed to do so.

What makes Rerum Novarum relevant today is not its specific prescriptions, many of which were products of their time. It is the evaluative framework underneath them, grounded in the principle that technology and economic systems must be judged by whether they protect human dignity and that the people most affected by those systems have claims that cannot be dissolved by efficiency arguments.

On May 25, 2026, Pope Leo XIV released his first encyclical, Magnifica Humanitas, on safeguarding the human person in the time of artificial intelligence [2]. The deliberate invocation of his predecessor's work, signed on the 135th anniversary of Rerum Novarum, was not a nostalgic gesture. The AI revolution,Leo XIV contends, poses the same category of question that industrialization posed: when a transformative technology reshapes labor, power, and human relationships, what do we owe each other, and what does it mean to remain human within it?

The encyclical's central evaluative question, drawn from John Paul II, is whether AI makes human life more worthy of the human person. That is a governance question in eight words, and one that applies as directly to a contracting shop or a supply chain organization as it does to a nation-state or a corporation. The encyclical is explicit that the question is not limited to regulation. It demands that the people who design, deploy, and use AI systems bear accountable responsibility for what those systems do to the humans inside them.

Critically, the encyclical does not treat AI as inherently good or bad. It treats technology as reflective of the intentions of those who devise, finance, regulate, and use it. That framing locates the moral variable not in the tool but in the governance decisions surrounding it, which is precisely where this piece locates it as well.

Leo XIV describes two paths using the biblical images of Babel andJerusalem. Babel represents the path where efficiency becomes the measure of everything, diversity of human judgment is flattened into uniformity, and the technology serves the ambitions of those who control it rather than the people it affects. What makes the Babel warning analytically specific is its diagnosis that the problem is not technological ambition itself but the choice of homogenization over communion, the sacrifice of human dignity for efficiency. Jerusalem represents something harder and slower, in which each person and institution contributes their section of the wall, rediscovering a common language not through uniformity but through harmony that comes from persons assuming their own role. The encyclical is unambiguous that the choice between these paths is made in the design choices of organizations, not just the policies of governments.

What the Research Is Telling Us

The CCM Institute's 2026 AI in Contracting report, drawing on 518practitioners across buy-side, sell-side, and dual roles in 16 sectors, offers a ground-level view of where the profession actually stands. The picture is more nuanced than either camp in my informal survey would predict, and in several respects more troubling than the headline enthusiasm numbers suggest.

Organizational enthusiasm for AI has risen sharply, from 36% in 2025to 56% in 2026. That is a significant single-year jump and reflects genuine institutional momentum. At the same time, personal practitioner enthusiasm has moderated, declining from 77% to 70%. The report interprets that moderation not as disengagement but as growing realism as professionals move from abstract possibilities to practical experience, encountering data limitations, governance gaps, and workload impacts along the way. The gap between what organizations are pushing for and what practitioners are experiencing in the field is itself a governance signal worth taking seriously.

The most telling single finding concerns over-reliance. When practitioners were asked how their concerns about specific AI risks had changed since implementation began, over-reliance on AI emerged as the most pronounced area of concern, with 40% indicating conditions had worsened. It was the only risk category where a majority reported deterioration. That is not a theoretical worry about future misuse. It is practitioners reporting what they are observing as AI tools become embedded in daily workflows, with human judgment and challenge being diluted rather than augmented.

The data on where practitioners see AI value reinforces the augmentation argument in contracting-specific terms. Risk assessment and compliance lead at 65%, contract performance monitoring at 60%, and contract generation at 55%. Negotiation support drops sharply to 29%. Practitioners are drawing a clear line between the procedural and the relational, between what AI can handle and where human judgment, context, and relationship management remain the irreplaceable inputs. That line tracks closely with what the CCMInstitute Benchmark report 2025 identified as the structural reality of contracting work, namely that most contracts remain static instruments, and when conditions change, organizations rely heavily on escalation, renegotiation, and informal workarounds that require exactly the kind of contextual judgment AI cannot supply.

The qualitative data from the survey adds texture that the percentages alone do not capture. Practitioners worry specifically about becoming input/output operators rather than strategic professionals. They worry about junior staff missing the foundational learning that handling basic tasks traditionally provided. And they flag a concern that is particularly relevant for senior leaders: non-contracting executives increasingly believeAI-generated contracts are complete and do not require professional review.That misconception is particularly dangerous when held at the leadership level because it drives decisions to bypass essential contract management processes entirely.

On job displacement, the data is more measured than the public discourse might suggest. Only 38% of CCM professionals express concern in varying degrees, notably lower than procurement or legal roles. The report attributes this partly to the nature of CCM work itself, which involves relationship building, strategic thinking, and complex problem-solving that requires human judgment and emotional intelligence. But the report's own framing of the decisive leadership question is worth stating directly: are we developing and using AI to reinforce disciplined commercial decision-making, or to bypassit? That question sits at the center of what this piece is arguing, and the fact that it emerges from practitioner research rather than theoretical analysis gives it weight.

The Integrity of the Transition

In work published with Tojin Eapen in Strategy and Leadership, we argued that organizations face a technology integrity challenge when adopting disruptive technologies, meaning the problem of remaining whole, coherent, and true to organizational identity while managing a transition that touches every dimension of how work gets done. We identified three facets of integrity that must be preserved simultaneously. Identity integrity is fidelity to what the organization actually is and does. Operational integrity is the continuity and coherence of how work is performed. Relational integrity is the trust-based relationships with employees, customers, partners, and stakeholders that give the organization its standing.

Those three dimensions map directly onto the governance questions that Magnifica Humanitas is raising. Identity integrity asks whether AI deployment changes what the organization fundamentally is in ways that were not chosen deliberately. For a contracting organization, the question is whether the shift to AI-assisted workflows preserves or erodes the judgment, expertise, and professional accountability that define the function. Operational integrity asks whether the new system actually works as designed in the full complexity of real conditions, not just in the pilot environment where it was validated.Relational integrity asks what the deployment does to the trust relationships that make the organization function, with employees whose roles change, with counterparts who now interact with AI-mediated outputs, and with the institutional relationships that carry institutional memory.

Organizations that shortcut the structured conversations these questions require tend to make AI deployment decisions that are technically sound in isolation but corrosive to the professional judgment, accountability structures, and relational fabric that the organization actually runs on. The two camps I encounter in client work are both ways of avoiding those conversations. One says the outcome is determined and positive; the other says it is determined and negative. Both relieve the organization of the responsibility to govern the transition deliberately.

The Augmentation Distinction

Augmentation and replacement are not points on a single spectrum. They represent fundamentally different organizational bets about where human value actually lives in a given workflow.

Consider a contract specialist in a defense acquisition shop. A significant portion of that professional's day involves retrieving and cross-referencing regulatory provisions, drafting standard clauses, formatting deliverable schedules, and logging correspondence. These are tasks where AI assistance is not only appropriate but overdue. AI tools that accelerate this work free the specialist to do more of what the role actually requires at its highest level, meaning negotiating terms under pressure, building relationships with counterparts, reading the room in a source selection, and knowing when a clause that is technically permissible is practically unenforceable. That is augmentation. The human is better at the job because the tool handles what the tool handles well. A recent WorldCC and Accenture report on adaptive contracting describes the human-machine division of labor in terms that align closely with the augmentation logic, with AI handling the monitoring, consolidation, risk modeling, and option-generating work, while humans retain responsibility for setting intent, interpreting consequence, navigating relationships, and governing trade-offs.[3]

Replacement logic works differently. It asks not what the AI can assist with, but what the human can be removed from. The answer, in the short term, is often more than organizations expect. First drafts of routine contracts. Initial supplier screening. Standard price analysis. A capable AI system can produce outputs in these areas that are good enough to pass initial review.

What replacement logic consistently underweights is the value embedded in the human doing the work. The contract specialist who drafts a first pass also builds institutional knowledge of that supplier relationship.The sourcing professional who runs the initial screening develops judgment about market conditions that no single transaction captures. When those humans are removed, the outputs may continue for a while but the institutional knowledge does not. Work, as the encyclical draws from John Paul II's Laborem Exercens, is not simply a means of generating income but a context for expression, relationships, and contributing to the community. Decisions that treat workers as removable from that context are not just operational decisions. They are decisions about what people lose when they are removed from meaningful participation in their profession.

The encyclical puts a finer point on the AI side of that equation than most operational analyses do. It notes that AI systems do not undergo experiences, do not possess a body, do not feel joy or pain, do not mature through relationships, and do not know from within what love, work, friendship, or responsibility mean. They can imitate language and analytical skills, but they do not understand what they produce. Their way of learning is statistical adaptation based on data and feedback, which can be very effective but does not imply inner growth. That is not a theological claim but an accurate description of how current AI systems function. The practical implication for organizations is that the judgment, relationship history, and contextual wisdom a human professional accumulates over a career cannot be replicated by a model trained on past transactions, however sophisticated that model becomes.

The cost of ignoring that distinction tends to surface in a dispute, a compliance failure, or a negotiation that goes wrong because no one in the room has the relationship history to read what is actually happening.

The Governance Question

The encyclical's argument is that the question is not what AI can do but what organizations, and the societies in which they operate, choose to do with what AI makes possible. That is a governance question, and one that too few organizations are engaging seriously.

Most AI strategy conversations I encounter focus on capability questions, meaning what the technology can do, what it costs, and how fast it can be implemented. The governance conversation, about who bears accountability for deployment decisions, what protections exist for affected workers, and how human dignity is preserved in redesigned workflows, tends to be deferred or treated as a compliance concern rather than a strategic one. Part 3 of theWorldCC adaptive contracting series notes that designing for adaptability requires explicit choices about which contractual terms are fixed, which are adjustable, and who, whether human, AI agent, or both in concert, is authorized to act.[4]

Magnifica Humanitas is direct on this point. It argues that when AI systems present themselves as neutral and objective, they end up reflecting and reinforcing the cultural assumptions and potential biases of their designers.It insists that entrusting an algorithm with the power to make consequential decisions without anyone bearing responsibility for that judgment is to handover the task of redefining the boundaries of human possibilities. And it goes further, arguing that calling for AI alignment with human values is insufficient if the ethical frameworks being aligned are determined by a small number of actors without broader accountability. This is not a singular person’s heterogenous conversation with faith and values. It is a conversation that impacts on those human agents who serve larger systems such as organizations and especially public systems that require them to be stewards of public resources. The encyclical calls for accountability at every stage, from those who design and develop AI systems to those who use them and rely on them for concrete decisions.

The encyclical also offers a useful reframe for leaders who worry that slowing down means falling behind. Calling for prudence, rigorous evaluation, and even at times a slower pace in adopting AI does not mean opposing progress. It is an exercise of responsible care for the people and institutions affected by those decisions. That reframe matters for the difficult conversations organizations need to have internally. Asking hard questions about AI deployment before committing to it is not obstruction. It is governance, and it is what the people in those organizations deserve from their leadership.

An AI deployment that reduces processing time by forty percent while systematically eliminating the human judgment that catches the error a model cannot detect has not produced a net gain. It has produced a system that is faster and more fragile, with accountability distributed so thinly that when something goes wrong, no one can identify where the decision was made.

What This Means for Contract, Supply Chain, and Operations Leaders

The practitioners and executives I advise are not theologians, and they should not need to be. But the framework that Magnifica Humanitas articulates, and that the technology integrity models that I have developed with colleagues, translates directly into questions that any serious leader can and should be asking before making AI deployment decisions.

The first question is what human value is embedded in a given task that does not appear in the output. The contract specialist's institutional knowledge, the sourcing professional's market judgment, and the operations manager's feel for supplier relationships are real and not always visible in the workflow diagram. Organizations that cannot answer this question before deployment are making a bet they have not consciously evaluated.

The second question is who is accountable when the AI-assisted output is wrong. If the answer is unclear, the deployment is not ready. Diffused accountability isa governance problem that technology is making easier to ignore. The encyclical's language on this is worth noting for any organization in a regulated environment, namely that the chain of responsibility must be identifiable and verifiable, and that those who design, authorize, and employAI systems must be held accountable for their decisions.

The third question is what happens to the people whose work changes. Augmentation and displacement both have workforce implications.The encyclical argues that every introduction of automation should be accompanied by verifiable measures to protect employment, retraining, and the participation of workers in the transition. Organizations that plan for this seriously, thinking about role redesign and the genuine value of experienced professionals, will make better decisions than those that treat headcount reduction as a natural byproduct of modernization.

The fourth question is whether the conversations are happening that need to happen. The technology integrity framework identifies three kinds of stakeholder dialogue that organizations consistently under invest in: conversations about what the organization actually is and wants to remain, conversations about how the transition will be managed operationally, and conversations about the trust relationships that the transition will affect. Organizations that avoid these conversations tend to find the underlying questions surfacing later as operational failures rather than strategic choices.

None of these questions require skepticism about AI's potential. The technology is real, the capabilities are substantial, and the organizations that learn to use it well will have meaningful advantages. But advantage is not the same as wisdom, and speed of adoption is not the same as quality of governance.

A Moment That Deserves More Than Binary Thinking

What strikes me about the release of Magnifica Humanitas is not that a pope has weighed in on artificial intelligence. It is that the most substantive recent argument against binary thinking about AI has come from an institution that most technology executives would not think to consult.

The encyclical does not tell organizations what AI to buy or how to implement it. What it does is insist that the evaluation criteria for those decisions include human dignity as a prior commitment, not an afterthought. It argues that technological progress without corresponding ethical and social progress produces a condition of having more without being more, and that in such a scenario individuals end up being evaluated principally according to the outcomes they produce rather than recognized as persons with inherent worth.That is not a religious claim but a governance claim, and one that the contract, supply chain, and operations professions are particularly well positioned to take seriously.

We are professionals who work in environments where accountability is contractual, where the allocation of risk is explicit, and where the question of who bears responsibility for an outcome is not rhetorical. Those instincts should be applied to AI deployment with the same rigor we apply to source selection or supply chain risk management.

The false binary, AI as salvation or AI as disease, is not just intellectually lazy. It is operationally dangerous because it relieves the people in the room of the responsibility to make careful decisions about what they are building and what they are willing to be accountable for.

The contracting profession and its organizations must decide if they want to take on a project of grandiose efficiency that flattens human diversity and accumulates power in the hands of those who control the technology, or whether it serves to take on the patient, shared, accountable work of building institutions and workflows that protect human dignity and distribute responsibility appropriately. The honest answer, for most organizations today, is that they have not yet asked the question clearly enough to know which they are building.

The question is not whether AI changes our professions. It will and it already has. The question is whether we govern that change deliberately, with clear eyes about what we value and what we owe each other, or whether we let the technology make those decisions by default.

That is a human question. It deserves a human answer.