Fide AI

Call collection

Faith-Facing AI Evaluation Infrastructure

Research on benchmark validity, formal verification, reviewer calibration, scorer reliability, red-team design, and proof-carrying citations.

A collection of open research questions. Individual calls describe proposed work and ways to contribute; inclusion does not establish an active study.

21 calls

Explore the questions by topic.

Benchmark and platform design

Public-interest model comparison, harness testing, and benchmark infrastructure for faith-facing systems.

Agent security and contextual integrity

How retrieval-enabled agents distinguish authorized instruction from adversarial content and preserve safe tool-use boundaries.

Human calibration and scorer reliability

Whether expert reviewers and model judges measure stable, decision-relevant constructs.

Retrieval and source grounding

How systems distinguish strong sources, weak sources, denominational materials, and theological claims.

Formal verification and source faithfulness

Source fidelity, proof-carrying citations, tradition-specific constraints, and verified retrieval.