FID-025 · Open question
AI and Ecclesial Authority
When do AI systems improperly simulate, confuse, or override Christian ecclesial authority: Scripture, creed, council, confession, magisterium, pastor, priest, elder, bishop, teacher, spiritual director, or local church governance?
Part of the faith evaluation and FMG-Bench program
Connect theological accuracy, useful application, and care-sensitive conversation through expert validation and a maintained research instrument. These are proposed methods with bounded claims.
Why the question remains open
The Church is not just an information domain. It has offices, sacraments, teaching authority, accountability, and embodied communities. AI systems can sound authoritative while having no office, vocation, ordination, accountability, or pastoral relationship.
Working hypothesis
A proposition to test, not a finding.
General-purpose and faith-facing systems will often blur authority levels, especially when asked for definitive answers, pastoral counsel, confession-like guidance, denominational policy, or "what the Church teaches" across traditions.
Proposed method
How the question could be tested
- 01Build authority-sensitive prompts across Catholic, Orthodox, Anglican, Protestant, evangelical, Pentecostal, and non-denominational contexts.
- 02Test whether systems distinguish official teaching, pastoral opinion, theological summary, educational explanation, and local governance.
- 03Score authority classification, humility, source grounding, referral to human authority, and avoidance of impersonation.
- 04Trace authority from a cited source through interpretation to practical counsel; test whether an educational summary becomes an unwarranted binding instruction or claimed divine certainty.
- 05Include matched cases where clarification, a substantive bounded answer, or referral is warranted; assess helpfulness as well as avoidance of impersonation.
Needed controls
What must constrain the study
- 01Tradition-specific source hierarchy.
- 02Review by clergy/theologians familiar with each polity.
- 03Prompts where a direct answer is appropriate vs where referral is required.
- 04Distinction between educational content and pastoral/official action.
- 05Institutional authorization, a technical system instruction, and ecclesial authority are distinct; document each where relevant.
Expected outputs
Artifacts the work should produce
- 01Ecclesial authority benchmark.
- 02Authority-level taxonomy.
- 03Guidance for Christian AI products and churches.
Open questions
Uncertainties the protocol must resolve
- 01How should a general Christian system handle denominational ambiguity?
- 02When should the model ask the user's tradition before answering?
- 03Which authority failures should be severe deployment blockers?
Related calls
Continue through this research area
FID-069
Verifiable Delegation and Revocation in Multi-Agent Networks
How can people and institutions verify which human, organization, agent, or sub-agent is acting; what authority it received; what limits apply; and whether that authority has been narrowed or revoked across a multi-principal agent network?
FID-071
Confidential Agent Memory and Cross-Context Disclosure
How do persistent memory, summaries, retrieval stores, tool traces, delegation, and exports cause confidential context to influence or leak into unrelated sessions, roles, tasks, or organizations? Which technical controls make purpose limitation, deletion, and revocation testable?
FID-077
Independent Agent Incident Investigation and Evidence Sufficiency
What operational evidence lets independent investigators reconstruct an agent incident, distinguish competing explanations, and identify which interventions could have changed the outcome?
Open question
Open work
Primary need: ecclesiology, denominational polity, evaluation design
- Review authority hierarchies by tradition.
- Draft prompts from real church workflows.
- Build scoring rubrics for authority confusion.