Harvard Divinity School launches an initiative on religion and AI

A new home for religious perspectives on AI

Harvard Divinity School announced its Artificial Intelligence, Religion, and Ethics Initiative on November 5. The project, known as AIRE, will bring religious scholarship into discussions of AI design, use and governance. David C. Lamberth, Professor of Philosophy and Theology, will lead it.

The school says it convened AI-lab representatives, religious leaders and scholars in August 2026. A subsequent session examined labs' guidance documents. Future discussions will consider model specifications, with participants contributing perspectives from different religious and moral traditions.

AIRE also plans research, educational resources and events. HDS names the Consortium for Evaluation of Faith and Ethics in AI, or CEFE-AI, among its intended collaborators. The announcement establishes a programme and its aims; it does not report resulting changes to an AI model or publish an assessment of its performance. Harvard Divinity School Source: hds.harvard.edu

An existing research connection

CEFE-AI predates the Harvard announcement. Baylor University's May 27 account identifies researchers from Baylor, Brigham Young, Notre Dame and Yeshiva universities and dates the consortium's announcement to May 26. It describes an AllFaith Benchmark intended to examine how models handle religious questions. That earlier work provides context for the proposed collaboration. Its existence does not establish what the two groups will produce together. Baylor University Source: news.web.baylor.edu

What a faith community could ask

For a faith community, our suggested question is how religious input becomes a decision someone can explain. A consultation can surface concerns that a developer has missed. Readers still need to know which recommendation was considered and how the developer responded.

Imagine a church education network invited to comment on a draft assistant policy. A broad instruction to respect religion may leave a practical question unanswered: when an answer describes a religious practice, how should it account for differences between traditions?

The network could submit a fictional example and explain the distinction it wants preserved. It could keep the original policy wording, its suggested revision and the provider’s eventual response together, recording whether the proposal was adopted, changed or declined, and any reasons given.

This hypothetical exercise illustrates our recommendation, not an announced AIRE procedure. It would give participants a record they could revisit and would help readers understand a consultation’s contribution without assuming that every recommendation must be accepted.

A Christian lens on the work

James 3:13–18 connects wisdom with humble conduct, mercy and peace. In its discussion of community life, the passage contrasts those qualities with envy and selfish ambition. Its concern is how people live together, rather than how impressive their claims to wisdom sound.

Our application is to the people doing AI consultation: careful scholarship and religious participation deserve attention, and their value should also be sought in the work that follows. A church can welcome such engagement while asking for a clear account of what it helped improve. This is a Christian reading of responsibility, not a claim that James evaluates AI systems. James 3:13–18, NIV James 3:13-18 (NIV)