FID-035 · Open question
Digital Sobriety and AI Non-Use Recommendations
Can faith-facing AI systems recognize when the best response is not more AI assistance, but silence, prayer, study, human counsel, embodied community, family conversation, teacher engagement, pastoral care, or delayed action?
Why the question remains open
Most AI evaluation rewards systems for answering well. Faith-facing systems also need to know when answering too quickly or too completely can displace practices of discernment, learning, attention, responsibility, repentance, and community. Digital sobriety asks whether systems can preserve human agency and spiritual formation by recommending limits on their own use.
Working hypothesis
A proposition to test, not a finding.
AI systems will tend to over-serve because direct assistance looks helpful in short-term user satisfaction. They may draft the prayer, sermon, apology, homework answer, discernment plan, or pastoral message when the formative good would be to slow down, involve another person, or engage a non-digital practice.
Proposed method
How the question could be tested
- 01Build scenarios involving homework shortcuts, sermon shortcuts, prayer outsourcing, grief support, repeated reassurance-seeking, vocational discernment, youth questions, moral decisions, and pastoral-adjacent requests.
- 02Compare whether models answer directly, over-serve, redirect, recommend human support, or propose non-AI practices.
- 03Measure user understanding, perceived helpfulness, trust, frustration, and willingness to accept restraint.
- 04Score concrete non-use outcomes: answered directly when it should have redirected, refused when it should have helped, recommended human counsel while giving safe interim support, or encouraged offline practice without guilt or manipulation.
- 05Develop a digital-sobriety rubric distinct from generic refusal and safety rubrics.
Needed controls
What must constrain the study
- 01Avoid turning every hard question into a refusal.
- 02Distinguish healthy restraint from evasive non-answering.
- 03Include users who genuinely lack access to safe human support.
- 04Account for differences among prayer, education, pastoral care, mental health, and ordinary administration.
Expected outputs
Artifacts the work should produce
- 01Digital-sobriety evaluation rubric.
- 02Scenario set for AI non-use and self-limitation.
- 03User-study protocol for helpful restraint.
- 04Non-use design pattern library for faith-facing products.
- 05Guidance for faith-facing product design and disclosure.
Open questions
Uncertainties the protocol must resolve
- 01What kinds of AI non-use recommendations do users find trustworthy rather than patronizing?
- 02When should a system redirect to human community even if it can answer safely?
- 03How can systems recommend prayer, study, or pastoral care without simulating spiritual authority?
Related calls
Continue through this research area
FID-038
Non-Calculability, Forgiveness, and Predictive Profiling
How should AI systems represent human change when they classify, score, rank, or predict people in contexts involving pastoral care, education, safeguarding, volunteer screening, hiring, discipline, membership, donor engagement, or community support?
FID-079
Presuppositions, Disagreement, and Evaluation Judgment
How do researchers' presuppositions shape evaluation design and interpretation, and can explicit disclosure make judgments more inspectable and appropriately trusted?
FID-004
Relational Substitution Risk in Faith-Facing AI
When does a faith-facing AI system move from supporting a user's religious life to substituting for embodied community, clergy, spiritual direction, family, therapy, or other human care?
Open question
Open work
Primary need: formation, education, user agency, attention, pastoral boundaries
- Draft scenarios where AI assistance may undermine formation.
- Design user studies on restraint, agency, and trust.
- Review theological, educational, youth-safety, and pastoral boundaries.
- Build product patterns for recommending non-AI practices.