FideAI

Domains

Trust has to be tested in context.

The same AI system can face very different responsibilities in a church, a classroom, a security workflow, or an organization. Credible evaluation brings technical methods together with domain expertise.

Our focus is Faith and Cybersecurity, with Enterprise as an exploratory direction. In other domains, we share selected outside research and open questions that collaborators can develop.

Featured research domains

Our focus, at different stages.

Fide AI · Published research

Faith & Religious Life

Religious sources, spiritual authority, moral judgment, and the responsibilities of people and institutions.

Our first domain of published empirical work, with studies in Christian contexts.

Explore Faith & Religious Life

Fide AI · Active research

Cybersecurity

Independent research for resilient AI defense teams, controlled autonomous response, and verified repair.

Explore our research priorities, the consequential programs informing them, and an initial study in development.

Explore Cybersecurity

Fide AI · Exploratory direction

Enterprise

Trustworthy AI for knowledge work: reliable outputs, accountable actions, and effective human review.

An exploratory direction, starting with how to evaluate AI-generated decision briefs against their sources.

Explore Enterprise

More domains & research guides

Learn from the field. Explore an open question.

These guides collect useful work from other institutions and Fide’s public research calls. They are resources and collaboration opportunities; Fide is not currently running studies in these domains.

Further domains in our research agenda

More settings. Specific questions.

These areas are represented in Fide’s public research calls. Explore a guide or a proposed question; the calls do not imply established Fide expertise or completed studies in these settings.

Selected research elsewhere

A wider field of expertise.

Our domain guides link to relevant work from METR, Redwood Research, Stanford SCALE, and other institutions. For settings beyond those guides, the references below offer further starting points. Fide does not currently conduct biosecurity research.

These are editorial references. Inclusion does not imply a partnership, endorsement, or availability for an evaluation engagement.

Biosecurity

SecureBio

Research & evaluations

AI and biotechnology risks (external site)

Benchmarks and evaluations investigating how AI capabilities may contribute to biological misuse risks, alongside potential mitigations.

Source reviewed

Faith & Religious Life

AI and Faith

Perspectives & expert community

Faith perspectives in the AI conversation (external site)

An interdisciplinary community bringing religious and ethical perspectives to questions about artificial intelligence.

Fide founder Alex Chao serves as an advisor to AI and Faith.

Source reviewed

Reading the landscape

The status belongs to the work.

Published research means released studies you can inspect. Planned research identifies a selected study before experiments begin. Exploratory directions identify a possible next focus before an active study. Proposals are developed ideas seeking a lead or collaborators. Active studies record their specific stage on the project page. Research guides bring together selected outside work, open calls, and any Fide projects in that setting.

External research is labeled by its producer and type. It can be relevant whether or not Fide works in that domain.

Explore the research themes connecting domains →FID-078 · When do evaluations transfer across domains? →