Seeking alignment: What religion can teach us about the past and future of AI
On stage from left to right: Panelists Syed Ishtiaque Ahmed, Suzanne van Geuns, Anna Su, Sarah Sharma and John Modern join in conversation to consider how the religious pasts of AI shape the foretelling of its future. Photography by Duane Cole.
When we consider a longer history of “religious” practices that train, prompt, or discipline human beings to live and think in “alignment” with higher powers and potent systems, what can we learn about artificial general intelligence?
What can past visions of religious alignment tell us about our visions for an AI future? And what might we learn from the people discarded and “corrected” in the pursuit of such success?
On April 8, 2026, as part of the 2025–2026 50th anniversary program of the Department for the Study of Religion, the Department, together with the Schwartz Reisman Institute for Technology and Society and the Data Sciences Institute hosted an interdisciplinary conversation to address questions on this topic.
Seeking Alignment: Religious Imaginaries in the Past and Future of AI brought together scholars from religion, media studies, computer science, and law to present their research and discuss how the religious pasts of AI shape the foretelling of its future.
AI as a modern religious revival
John Modern, who teaches in the Department of Religious Studies at Franklin & Marshall College, opened by challenging the idea that AI is simply something we “use”, describing contemporary neural networks as part of a broader neuromatic paradigm, where human and machine intelligences loop into one another.
In this framework, AI doesn’t just perform tasks, it rearranges how value, attention, and authority circulate. The result is a mix of techno‑optimism, dread, and faith that often mirrors religious devotion, complete with prophets, rituals, and apocalyptic fears.
“AI systems in their massive complexity and capacities for surveillance and prediction approach the criteria that humans have long used to imagine divinity.” said Modern, calling for the humanities and the study of religion to look past the technical bafflement of AI and aggressively critique its cultural machinery.
“These systems are not divine…They are rather discursive. They are utterly human. They are systems for probabilistic mimicry that are ever improving their odds of imitating us as humans in turn imitate their imitators.”
John Modern warns that we are treating AI with a form of religious zealotry, which creates a dangerous feedback loop that threatens to flatten human thought and culture.
The illusion of “the tool”
Sarah Sharma, a professor of Media Theory and Director of the Institute of Communication, Culture, Information and Technology at the University of Toronto, extended this critique through a feminist lens, pushing back against the familiar framing of “AI as tool.”
Sharma argued that the design and power structures of modern technology, including big tech and generative AI, are not neutral, but are driven by a patriarchal logic rooted in control that treats technologies - and the people who sustain them - as disposable instruments.
“The big tech patriarch doesn't really think of technologies as tools but rather as environments in which to cultivate, manage, and extend their power,” said Sharma, illustrating that the secular religion of Big Tech is actually driven by ancient theological impulses: the desire to play God, the fear of mortality, and the aspirations to reaching a technology-driven utopia. Sharma’s new book Insufferable Tools: Feminism Against Big Tech expands directly on the ideas she outlined in her talk offering a sharp critique of how tech giants weaponize media theory to preserve patriarchal control.
Sarah Sharma illustrates how Big Tech companies establish an environment of artificial, frictionless abundance, mimicing a maternal paradise where all physical needs are instantly met, requiring no reciprocal human connection.
Machine-like thinking
AI alignment was then reframed by Suzanne van Geuns, a professor of Religion and Technology at the University of Wisconsin, as a religious and disciplinary project. Rather than looking at AI as a secular replacement for traditional religion, van Geuns illustrated the concept of “computational creep”—a phenomenon where humans increasingly attempt to optimize and better themselves by thinking like machines rather than vice versa.
“When we write first principles for systems that are not capable of empirical experience or trial and error, it might seem like we really are the gods…” said van Geuns, “But computational creep suggests that we are actually already thinking more like the systems we want to teach how to think.”
Suzanne van Geuns illustrates how Silicon Valley’s obsession with “AI value alignment” relies on the exact same religious interpretation methods used by conservative, anti-feminist "biblical womanhood" bloggers.
Silencing difference in AI alignment
Syed Ishtiaque Ahmed, an assistant professor of computer science at the University of Toronto and SRI faculty affiliate, illustrated that AI value alignment is not a neutral, high-level philosophical quest, but a highly disciplined “digital factory floor” that actively enforces a Western-centric, secular worldview on a diverse global society.
Examining data annotation work in Bangladesh, Ahmed revealed how Western secular guidelines erase religious meaning and enforce artificial consensus. Annotators are trained to suppress their own ethical frameworks, exposing how alignment relies on silencing difference at the very point where AI is made.
Anna Su, Suzanne van Geuns, and Sarah Sharma in discussion during the panel portion of the event.
Rehearsing a familiar theological drama
Professor of Law and SRI Research Lead Anna Su connected these concerns to law and human dignity. AI safety discourse, she noted, often relies on apocalyptic narratives, from runaway superintelligence to salvation through machine enlightenment. These stories echo older attempts to grapple with overwhelming power. Just as constitutional law sought to constrain kings and states, emerging debates about legal alignment ask how AI might be bound by democratically legitimate rules.
“Modernity imagine itself as having left religion behind. But what actually happened is that the theological categories got translated into secular ones stripped of their origins but not its logic. And so AI alignment is the latest instance of this. When we talk about safe models and human values encoded in machine behavior, we are rehearsing a drama that theology has grappled with over the centuries.”
The panel closed with a shared warning: treating AI as inevitable shrinks the space for refusal, responsibility, and alternative futures. Rather than surrender alignment to a technical elite, the speakers called for plural ethics, public accountability, and sustained critical engagement.
Want to learn more?
Browse stories by tag:
- AI
- AI safety
- AI trust
- Computer science
- Computer security
- Copyright
- Cybersecurity
- Data
- Democracy
- Economics
- Education
- Engineering
- Ethics
- GPO-AI
- Governance
- Health
- Human Rights
- In the Media
- Jobs
- LLMs
- Law
- Normativity
- Philosophy
- Political Science
- Privacy
- Privacy Series
- Psychology
- Public Policy
- Recommenders
- Regulation
- Religion
- Reports
- Trust
- Workshops
