What the Pope Can Teach Us About Agentic AI

Kevin O'Marah

May 28, 2026 • 5 min read

---

Pope Leo XIV this week published a 42,000-word encyclical (aka, position paper) on AI. His message intentionally echoed a similar encyclical released by Pope Leo XIII 135 years ago on the rising Industrial Revolution. The common thread is a call for human dignity in work.

Now, as then, the ethics of power, wealth, and fairness are topical both as a matter of justice and scalability. 

Papal Wisdom Aligns with Good Governance for Agentic Supply Chains

For supply chain leaders, the AI push focuses on agentic systems. Zero100 research on the workflow-defined organization models we call PowerThreads points to massive productivity gains (30-50%). But only if we keep humans in the loop to tune and scale AI’s learning capabilities.  Efficiency is great, but automation without a human touch seems increasingly tone deaf. 

Plus, this efficiency comes with new risks. COOs we surveyed in Q4 2025 rated cybersecurity as the top overall business continuity risk for 2026. Perhaps this is because AI agents increase the attack surface for bad actors with novel ways to do harm. A recent McKinsey study cites agents breaking policy, failing decision logic, and developing divergent behaviors as emerging areas of concern. 

Summarizing the Pope’s key points yields a solid short list of leadership tenets for managing both the upside and downsides of agentic systems in supply chain operations: 

• Keep a named human accountable. If an agent approves, denies, escalates, sanctions, or reallocates something important, have a clear human owner. 
• Tier agentic systems by autonomy and impact. Routine drafting, monitoring, and analysis can be more automated than decisions affecting employment, safety, supplier treatment, or customer rights. 
• Audit for fairness, not just model accuracy. Monitor AI for bias, opacity, and arbitrariness as customers, suppliers, and employees increasingly deal with agents instead of people.
• Design work for agentic systems. As agents change roles and compress management layers, leaders should treat job and organization design as a paramount duty. 

Cognitive Atrophy and the Meaning of Work with AI 

MIT published a recent study measuring how essayists’ brains functioned differently with LLM support as compared to brain-only work. After four months, the takeaway was scary – LLM users “consistently underperformed at neural, linguistic, and behavioral levels.” AI makes us more productive, but without stretching our capabilities, it risks turning us into automatons.  

The Pope recognizes this when he urges that AI be optimized for human dignity, not just efficiency. For supply chain leaders, this is not only a matter of morality; it is a reminder that we are rarely locking in hard automation so much as using talented operators to train a network of agents and people. The goal is a perpetual digital kaizen that amplifies rather than replaces human capabilities. 

Toyota North America embraced this approach in a digital transformation of its planning process. The monthly cycle had been a stressful, spreadsheet-based process, straining the morale of 40-50 planning people. The new process lets planners build and consider various scenarios to help choose better plans, while reducing work hours by 60-80%. Jason Ballard, VP, Digital Innovations for Toyota, in an interview with Deloitte, describes the experience for planners as a victory:  

“…now their work is uplifted. They’re taking on a much wider scope of responsibility. They’re getting stretched. They’re learning new parts of the operations. For the others, it’s allowing them to tap into their career aspirations.” 

Work, after all, can be a source of meaning for people and society, if technology serves to elevate, not eliminate, the job. 

History’s Lesson for Scaling AI 

Pope Leo XIII in 1891 warned about industrialization’s threat to human dignity:  

“…now their work is uplifted. They’re taking on a much wider scope of responsibility. They’re getting stretched. They’re learning new parts of the operations. For the others, it’s allowing them to tap into their career aspirations.” 

The following five decades scaling industrial technology saw massive gains in productivity, but also terrible wars, revolutions, and depressions. 

Politicians, financiers, and technologists now, as then, race to dominate the new technology, with little incentive to pump the brakes. Is it a coincidence that in 2026, as in 1891, operations leaders have the critical role of tempering its power?

I don’t think so.