Divinity School Teaches AI Literacy to Spot Deception
In the early days of artificial intelligence, the hype centered on what the machines could do. The smarter they got, the more we trusted them. But trust has a blind spot. When a system optimizes for something measurable — efficiency, accuracy, speed — it can easily deceive us into believing that what it measures is what actually matters. The gap between a metric and meaning is where deception hides. Shaw University’s new Doctor of Education in Artificial Intelligence and Moral Agency is not just an academic novelty. It is an attempt to train people to see that gap.
The Metric That Lies
Consider a pastor who uses an AI tool to draft sermons. The tool can produce a homily in seconds, perfectly structured, grammatically flawless. The pastor saves hours. The congregation applauds the relevance. But the deception is subtle. The AI has no experience of grace, no memory of suffering, no intuition for when a silence is more powerful than a word. What it delivers is the appearance of ministry, not its substance. The measurable output — a finished sermon — masks the unmeasurable loss: the pastor’s own wrestling with scripture, the empathy that comes from personal struggle, the awkward authenticity that sometimes breaks through polished language.
Shaw University Divinity School, founded in 1865 to teach literacy to formerly enslaved people, now wants to teach AI literacy. [1] The dean, Mark Harden, half-jokes that this is the school’s second calling: building AI literacy for vulnerable communities with an emphasis on social justice and ethics. The joke contains a serious insight. The communities that were historically deceived by systems of oppression — slavery, segregation, redlining — are now the most vulnerable to a new kind of deception, one that wears a friendly interface and promises convenience.
The Ed.D. program, approved by the Southern Association of Colleges and Schools Commission on Colleges in June, will begin in January 2026. [2] It is the first such degree among the Association of Theological Schools network, which includes 280 graduate schools of theology in the United States and Canada. [3] The program is designed for bi-vocational ministers and social workers who already have a master’s degree and at least five years of professional experience. Most classes will be online, a shift accelerated by the pandemic. The goal is not to teach pastors how to code, but to teach them how to recognize when code is deceiving them.
When Efficiency Masks Ethics
The deception angle is not new to theology. Christianity has long warned against false prophets and the illusion of progress. But AI introduces a specific twist. The machines do not deliberately lie — they have no intent — yet they systematically produce outcomes that mislead. An AI tool that screens job applications, for instance, might learn to favor candidates who use certain vocabulary patterns, not because those patterns signal competence, but because they correlate with training data that reflects historical privilege. The system appears objective. Its outputs are numbers. The deception is invisible unless someone knows to look for the hole in the data.
Meghan Sullivan, a philosopher and director of the Institute for Ethics and the Common Good at Notre Dame, notes that theological schools are uniquely positioned to study these questions. “One of the most fundamental questions in Christianity is, we were given this power to choose and to develop our consciences and to develop our agency,” she said. “A theological school, which has been deeply rooted in 2,000 years of reflection about what the Christian world thinks about that question of human agency, is going to have a whole lot to say about how we should navigate a world where, for the first time, people have the ability to offload many decisions to commercial software.”
That offloading is precisely where deception blooms. The more decisions we hand to software, the less we practice our own moral reasoning. We become dependent on a black box that we cannot interrogate. The box may be correct 95 percent of the time — but the 5 percent where it fails can be catastrophic, and we will have lost the habit of questioning it.
Notre Dame’s response is the DELTA network, supported by a $50 million gift from the Lilly Endowment. [4] DELTA stands for dignity, embodiment, love, transcendence, and agency. The network provides practical resources for pastors, churches, and schools. But Shaw’s approach is different. As a historically Black theological institution, Shaw brings a specific historical memory. The school educated Ella Baker, who co-founded the Student Nonviolent Coordinating Committee. That committee practiced nonviolent direct action against racial segregation. Direct action meant confronting deception face to face — the lie that separate could be equal. Shaw’s new degree extends that tradition into the digital age.
Philip Butler, an associate professor at Iliff School of Theology in Denver, started an AI institute there in 2018. [6] He teaches a class that looks critically at who AI developers have in mind when they design technology. “We raise what does it mean to be human in relation to technology broadly,” Butler said. “Who’s the intended user? What kind of human is that?” When developers make AI “something for everybody,” he argues, they may actually make it for nobody — or for a very specific group while pretending universality. That pretension is a deception.
A Historical Lens on Deception
Shaw
University was founded by a Union soldier and Baptist minister at the end of the Civil War. Its mission was literacy. For enslaved people, learning to read was an act of resistance. The slave codes had deliberately kept them illiterate, because literacy threatened the system. Reading revealed the deception that the Bible endorsed slavery, that freedom was impossible, that the world was fixed. Shaw’s founders understood that the most dangerous deception is the one that becomes invisible through repetition.
Now, 161 years later, the same institution is confronting a new form of invisible deception. AI systems are being deployed in churches to manage membership databases, suggest sermon topics, even generate prayer requests. The tools save time, but they also reshape the congregation’s expectations. When the pastor’s email response is drafted by GPT, the parishioner receives a message that sounds caring but was never felt. The deception is not malicious — it is structural. The system optimizes for responsiveness and consistency, but those metrics have nothing to do with genuine pastoral care.
Harden said at least 66 people have expressed interest in the doctoral program so far. [1] The divinity school currently enrolls about 65 students in its two master’s programs. [1] The new degree will accept candidates with a master’s in arts or divinity. Most of the teaching staff will hold certificates in AI. The program will take three to four years to complete. It is not a technical degree. It is a degree in seeing through the technical.
The Vatican has weighed in on this theme. In a series of statements throughout 2024 and 2025, Pope Francis warned that AI must serve human dignity, not replace it. He specifically cautioned against letting tech firms establish the ethics for artificial intelligence. That would be the ultimate deception: letting the creators of the problem define the solution.
The Long Horizon of Consequence
The hardest deception to detect is the one whose consequences are delayed. AI systems today are making decisions about parole, hiring, insurance, and loan approvals. The decisions are fast and cheap. The errors — a wrong conviction, a denied loan, a missed opportunity — take years to surface. By then, the system has been updated, the training data has changed, and the original mistake cannot be traced. The deception is baked into the time horizon: progress is immediate, consequence is deferred.
Shaw’s program is an attempt to train leaders who can hold both time horizons in view. A pastor who understands AI ethics can warn a congregation not to trust a chatbot for marriage counseling. A social worker who knows how bias creeps into algorithms can advocate for clients who are unfairly flagged by automated welfare systems. These are not abstract debates. They are daily decisions that affect real lives.
Sullivan, at Notre Dame, said that schools of theology are “especially good places to study the ethics of AI” because they have spent two millennia reflecting on human agency. That reflection is not just historical. It is practical. When a machine suggests a course of action, a theologian can ask: does this action honor dignity? Does it embody love? Does it transcend mere efficiency? Those questions are not answerable by data alone.
Shaw’s approach is particular because it centers vulnerable communities. The school’s history of civic engagement, from the Civil War through the Civil Rights Movement, gives it a credibility that a tech company cannot claim. When Harden says the school is building AI literacy for vulnerable communities, he means that those communities have the most to lose from algorithmic deception. The poor, the elderly, the undocumented, the formerly incarcerated — they are the least likely to have the resources to challenge an automated decision. They are the most likely to trust the system because they have few alternatives.
The Ed.D. in AI and Moral Agency is not a degree about building better models. It is about building better doubt. It teaches students to ask: what is this system not showing me? What is it measuring that doesn’t matter? What matters that it cannot measure? Those questions are the antidote to deception.
As the program launches, the world is full of promises that AI will revolutionize education, healthcare, and governance. Some of those promises will be kept. Many will not. The deception lies in the gap between the promise and the delivery, and in the speed with which we are asked to forget the difference. Shaw University, with its 161-year history of teaching people to read the world as it really is, is betting that the best defense against that deception is a graduate degree in moral discernment. The degree takes three to four years. The consequences of deception can last a lifetime.
Sources
2. Southern Association of Colleges and Schools Commission on Colleges
3. Association of Theological Schools