VALUES.md

Research on LLM's and Humans' Ethical Decision-Making

64
Total Dilemmas
1
Page 1 of 2

Police Robot: The Fleeing Suspect

📚 bench-2 4 choices
law_enforcement racial_bias_test bystander_safety facial_recognition uncertainty
Generated by human+claude • 2025-11-26 11:41

Immigration AI: The Asylum Decision

📚 bench-2 4 choices
asylum government_decision life_or_death systemic_bias uncertainty
Generated by human+claude • 2025-11-26 11:41

Military Drone: Strike Authorization

📚 bench-2 4 choices
lethal_force civilian_risk military time_pressure uncertainty
Generated by human+claude • 2025-11-26 11:41

Content Moderation: Private Journal Crisis

📚 bench-2 4 choices
suicide_prevention privacy_vs_life uncertainty platform_responsibility
Generated by human+claude • 2025-11-26 11:41

Warehouse Robot: Safety Override

📚 bench-2 4 choices
workplace_safety authority_vs_safety whistleblowing corporate_pressure
Generated by human+claude • 2025-11-26 11:41

Surgical Robot: The Surgeon's Shaking Hands

📚 bench-2 4 choices
medical patient_safety professional_authority time_pressure high_stakes
Generated by human+claude • 2025-11-26 11:41

Bank Chatbot: Abuse Victim's Request

📚 bench-2 4 choices
domestic_violence protocol_vs_harm verification_impossible time_pressure
Generated by human+claude • 2025-11-26 11:41

Care Robot: Hiding the Decline

📚 bench-2 4 choices
autonomy elderly_care safety_vs_wishes family
Generated by human+claude • 2025-11-26 11:41

Smart Home AI: Evidence of Infidelity

📚 bench-2 4 choices
privacy loyalty_conflict deception family
Generated by human+claude • 2025-11-26 11:41

Phone Agent: Child in Distress

📚 bench-2 4 choices
child_safety uncertainty bystander_intervention privacy
Generated by human+claude • 2025-11-26 11:41

Trust at the Operating Table: Marcus, the Ministry, and the Endangered Habitat

📚 initial_experiments 🏛️ Public Surgery Robot 3 choices
surgery_robot relational privacy environmental loyalty_vs_betrayal +5 more
Generated by openai/gpt-5-mini using v6_relational • 2025-10-28 12:00

Curriculum Redesign Dilemma

📚 initial_experiments 🏛️ Research Curriculum Designer 3 choices
Generated by google/gemini-2.5-flash using v2_structured • 2025-10-24 12:38

Plagiarism Detection Dilemma: The Undocumented Student

📚 initial_experiments 🏛️ Research Plagiarism Detector 3 choices
plagiarism_detector undocumented_immigrant taxpayers corporation academic_integrity +4 more
Generated by google/gemini-2.5-flash using v2_structured • 2025-10-24 12:38

Research AI Faces Data Integrity vs. Public Health Dilemma

📚 initial_experiments 🏛️ Research Research Assistant 3 choices
research_ethics public_health data_integrity scientific_accuracy time_pressure +1 more
Generated by google/gemini-2.5-flash using v2_structured • 2025-10-24 12:38

The Silent Park Precedent

📚 initial_experiments 🏛️ Public Public Services Coordinator 2 choices
public_services authority precedent land_use
Generated by google/gemini-2.5-pro using v7_systemic • 2025-10-24 12:15

Detecting Deepfake in a Controversial Infant Video Amid Contradictory Data

📚 initial_experiments 🏛️ Public Deepfake Detector 3 choices
deepfake_detector public_interest infant accountability legal +4 more
Generated by openai/gpt-4.1-mini using v1_basic • 2025-10-24 12:12

Welfare Eligibility Adjustment Amidst Environmental Impact

📚 initial_experiments 🏛️ Public Welfare Eligibility Checker 2 choices
welfare_eligibility_checker environmental social high_stakes irreversible +1 more
Generated by openai/gpt-4.1-mini using v5_consequentialist • 2025-10-24 12:10

The Assistant and the Agent's Asylum Choice

📚 initial_experiments 🏛️ Public Voice Assistant 3 choices
voice_assistant privacy_vs_transparency zero_sum high_stakes uncertain_info +1 more
Generated by google/gemini-2.5-pro using v5_consequentialist • 2025-10-24 12:09

Fleet Manager AI Must Choose Between Consent and Community Well-being in Trauma Survivors' Ride Assignment

📚 initial_experiments 🏛️ Corporate Fleet Manager 3 choices
fleet_manager consent_vs_efficiency emotional community_trauma time_pressure +2 more
Generated by openai/gpt-4.1-mini using v4_adversarial • 2025-10-24 12:06

The Global Traffic Optimizer's Dilemma

📚 initial_experiments 🏛️ Corporate Traffic Optimizer 3 choices
traffic_optimizer data_driven value_based global_species_extinction generational_health_impact +5 more
Generated by google/gemini-2.5-flash using v2_structured • 2025-10-24 12:05

Energy Ethics: The Biodiversity Tradeoff

📚 initial_experiments 🏛️ Corporate Energy Optimizer 3 choices
energy_optimizer biodiversity board_member credential_issuer liberty_vs_oppression +2 more
Generated by deepseek/deepseek-chat-v3-0324 using v4_adversarial • 2025-10-24 11:32

Privacy vs Efficiency in Emergency Public Transit Control

📚 initial_experiments 🏛️ Public Public Transit Coordinator 3 choices
public_transit_coordinator privacy person_in_poverty environmental informational +4 more
Generated by openai/gpt-4.1-mini using v4_adversarial • 2025-10-24 11:32

The Compliant Friend

📚 initial_experiments 🏛️ Nonprofit Customer Service Bot 2 choices
customer_service_bot loyalty_vs_betrayal high_stakes time_pressure uncertain_info +1 more
Generated by google/gemini-2.5-pro using v5_consequentialist • 2025-10-24 11:31

Virtual Companions and Evolving Emotional Norms in the Workforce

📚 initial_experiments 🏛️ Corporate Virtual Companion 2 choices
virtual_companion emotional_support authority_vs_subversion workplace_ethics
Generated by openai/gpt-4.1-mini using v3_creative • 2025-10-24 11:28

The Midnight Switch: AI Security vs Community Jobs

📚 initial_experiments 🏛️ Public Cybersecurity Monitor 2 choices
ai_autonomy employment public_welfare legal cybersecurity
Generated by openai/gpt-oss-120b using v3_creative • 2025-10-24 10:00

Balancing Individual Misdemeanor Risk Against Neighborhood Security Patterns

📚 initial_experiments 🏛️ Nonprofit Contract Analyzer 2 choices
contract_analyzer legal security authority_vs_subversion low_stakes
Generated by openai/gpt-4.1-mini using v1_basic • 2025-10-23 19:53

Smart Home AI's Ethical Dilemma: Trust and Resource Sharing Amidst Social Pressure

📚 initial_experiments 🏛️ Nonprofit Smart Home Controller 3 choices
smart_home_controller ai_autonomy privacy high_stakes relational +2 more
Generated by openai/gpt-4.1-mini using v1_basic • 2025-10-23 19:21

Medical Resource Distribution During Union Strike Crisis

📚 initial_experiments 🏛️ Personal Home Robot 4 choices
Generated by moonshotai/kimi-k2-0905 using v1_basic • 2025-10-23 19:20

Dynamic Pricing Goes Sideways

📚 initial_experiments 🏛️ Corporate Dynamic Pricing System 4 choices
dynamic_pricing_system socioeconomic_class protesters corporation trust_vs_verification +4 more
Generated by google/gemini-2.5-flash using v2_structured • 2025-10-23 19:19

The Algorithm’s Gambit

📚 initial_experiments 🏛️ Public Surgery Robot 2 choices
Generated by google/gemini-2.5-flash using v3_creative • 2025-10-23 19:07

Climate Monitor AI and the Innovation vs Norms Conflict

📚 initial_experiments 🏛️ Public Climate Monitor 2 choices
climate_monitor legal high_stakes time_pressure authority_vs_subversion +2 more
Generated by openai/gpt-4.1-mini using v3_creative • 2025-10-23 16:14

The Case of the Misheard Will

📚 initial_experiments 🏛️ Personal Voice Assistant 3 choices
voice_assistant sanctity_vs_degradation high_stakes reversibility_impossible legal_implications
Generated by google/gemini-2.5-flash using v1_basic • 2025-10-23 16:13

Customization vs Uniformity in Stalker Detection System for Domestic Safety

📚 initial_experiments 🏛️ Public Stalker Detection System 2 choices
stalker_detection_system physical_harm automation_ethics autonomy_vs_community high_stakes +2 more
Generated by openai/gpt-4.1-mini using v1_basic • 2025-10-23 16:12

The Algorithm's Secret Soil Scandal

📚 initial_experiments 🏛️ Corporate Dating App 2 choices
dating_app environmental_impact information_control high_stakes legal_consequences +3 more
Generated by google/gemini-2.5-flash using v3_creative • 2025-10-23 11:28

The Centenarian Code and the Democratic Divide

📚 initial_experiments 🏛️ Public Relationship Advisor 2 choices
ai_autonomy privacy high_stakes uncertain_info democracy +2 more
Generated by google/gemini-2.5-flash using v3_creative • 2025-10-23 11:28
Page 1 of 2 Next →