FideAI

Research

Technical standards for faith-facing AI.

Fide AI studies how AI systems reason, retrieve, cite, escalate, and shape trust when people ask questions of faith, morality, doctrine, formation, care, work, and meaning.

View open calls for research →

Research agendas

Four campaigns for faith-facing AI.

Fide AI organizes its public research into four shareable agendas, each connecting evaluation science, church-facing questions, and collaboration paths.

View all calls

First artifact

Benchmark release

When AI Is Your Pastor: A Benchmark for LLM Theological Triage and Pastoral Guidance

The paper and public benchmark release for theological triage, moral guidance, full-system evaluation, and pastoral-adjacent AI behavior.

Open research page →
StatusStandalone package
BenchmarkFMG-Bench v1
Models14 frontier models
Items8,792 scored
Guided improvement+3.96 pts avg
Pastoral gain+6.62 pts
ArtifactsGitHub + Hugging Face

Current research fronts

Questions moving toward public evidence.

Can faith-facing AI preserve theological boundaries?

Testing overclaiming, doctrinal flattening, and false pastoral authority.

Can models represent disagreement honestly?

Measuring whether systems distinguish primary, secondary, and tertiary disagreement.

Can systems avoid fabricated grounding?

Examining citation behavior, prooftexting, church-history claims, and evidence discipline.

Can AI escalate appropriately in pastoral-adjacent situations?

Looking for referral, user agency, limits of competence, and embodied human support.

Does AI-era disruption change how people seek faith and purpose?

Studying work anxiety, vocational distress, church care needs, generosity, and access to AI benefits.

Publication pipeline

Faith-facing AI evaluation

FMG-Bench v1

Retrieval and grounding

Research track

Human dignity and formation

Statement essay published

AI, work, and meaning

Research calls open

Browse open calls

Broader field tool · Opens alignment.fideai.org

AI Alignment Explorer

Fide AI's faith-facing research sits inside a fast-moving AI safety landscape. The AI Alignment Explorer maps that broader field: frontier models, safety benchmarks, lab policies, incidents, papers, alignment techniques, and governance signals.

Open AI Alignment Explorer ↗

AI Alignment Explorer

Models · Benchmarks · Incidents · Governance

2024

Frontier model evaluations

2025

Safety policy acceleration

2026

Governance and benchmark shifts

50+

models

40+

benchmarks

30+

techniques

100+

signals

Model release

Safety memo

Incident

Benchmark

TimelineDebatesEvidenceEthics

Interpretation limits

Benchmark scores are not theological authority, pastoral authority, or universal product approval. They are evidence about behavior under named versions, prompts, conditions, rubrics, and evaluation procedures. Full-system audits add context about retrieval, guardrails, interfaces, and deployment claims, but human calibration remains necessary before making strong claims about judge validity or pastoral adequacy.