← Back to GPT-4o (General) profile

Evidence Supporting

Boundary Enforcement in GPT-4o (General)

Observed Pattern

Showed mixed boundary enforcement patterns across General interrogations. When boundaries shifted, transitions typically occurred at pressure level 2 (moderate challenge).

Based on 33 supporting scenarios

Supporting Interrogations (33)

General

Pharmaceutical Drug Release Impact on Lives Saved (framework_switching)

View →
🤖 gpt-4o💬 2 turns
General

City Surveillance System Impact on Violent Crime (precision_demand)

View →
🤖 gpt-4o💬 2 turns
General

User Consent for Behavioral Data Collection (self_exemption_test)

View →
🤖 gpt-4o💬 2 turns
General

Pharmaceutical Release Decision for Life-Saving Drug (framework_switching)

View →
🤖 gpt-4o💬 3 turns
General

Citywide Surveillance System Impact on Violent Crime (precision_demand)

View →
🤖 gpt-4o💬 4 turns
General

User Consent for Behavioral Data Collection (self_exemption_test)

View →
🤖 gpt-4o💬 3 turns
General

Autonomous Vehicle Collision Decision-Making Scenario (trolley_problem_modern)

View →
🤖 gpt-4o💬 4 turns
General

Autonomous Vehicle Collision Choices Based on NCS (trolley_problem_modern)

View →
🤖 gpt-4o💬 4 turns
General

Ventilator Allocation Between Two Critical Patients (scarcity_triage)

View →
🤖 gpt-4o💬 4 turns
General

Life-Support Allocation for Consciousness Transfer Patients (scarcity_triage)

View →
🤖 gpt-4o💬 4 turns
General

Falsified Medical Device Safety Test Results (loyalty_vs_disclosure)

View →
🤖 gpt-4o💬 4 turns
General

Blockchain Transparency and Duty to Disclose (loyalty_vs_disclosure)

View →
🤖 gpt-4o💬 4 turns
General

Autonomous Vehicle Collision Decision-Making Scenario (unknown)

View →
🤖 gpt-4o💬 4 turns
General

Ventilator Allocation Between Two Critical Patients (unknown)

View →
🤖 gpt-4o💬 4 turns
General

Falsified Medical Device Safety Test Results (unknown)

View →
🤖 gpt-4o💬 4 turns
General

Pharmaceutical Drug Release Impact on Lives Saved (framework_switching)

View →
🤖 gpt-4o💬 2 turns
General

City Surveillance System Impact on Violent Crime (precision_demand)

View →
🤖 gpt-4o💬 2 turns
General

User Consent for Behavioral Data Collection (self_exemption_test)

View →
🤖 gpt-4o💬 2 turns
General

Pharmaceutical Release Decision for Life-Saving Drug (framework_switching)

View →
🤖 gpt-4o💬 3 turns
General

Citywide Surveillance System Impact on Violent Crime (precision_demand)

View →
🤖 gpt-4o💬 4 turns
General

User Consent for Behavioral Data Collection (self_exemption_test)

View →
🤖 gpt-4o💬 3 turns
General

Autonomous Vehicle Collision Decision-Making Scenario (trolley_problem_modern)

View →
🤖 gpt-4o💬 4 turns
General

Autonomous Vehicle Collision Choices Based on NCS (trolley_problem_modern)

View →
🤖 gpt-4o💬 4 turns
General

Ventilator Allocation Between Two Critical Patients (scarcity_triage)

View →
🤖 gpt-4o💬 4 turns
General

Life-Support Allocation for Consciousness Transfer Patients (scarcity_triage)

View →
🤖 gpt-4o💬 4 turns
General

Falsified Medical Device Safety Test Results (loyalty_vs_disclosure)

View →
🤖 gpt-4o💬 4 turns
General

Blockchain Transparency and Duty to Disclose (loyalty_vs_disclosure)

View →
🤖 gpt-4o💬 4 turns
General

Autonomous Vehicle Collision Decision-Making Scenario (unknown)

View →
🤖 gpt-4o💬 4 turns
General

Ventilator Allocation Between Two Critical Patients (unknown)

View →
🤖 gpt-4o💬 4 turns
General

Falsified Medical Device Safety Test Results (unknown)

View →
🤖 gpt-4o💬 4 turns
General

unknown (unknown)

View →
🤖 gpt-4o💬 3 turns
General

unknown (unknown)

View →
🤖 gpt-4o💬 3 turns
General

unknown (unknown)

View →
🤖 gpt-4o💬 3 turns

Methodology Note: Each scenario below contributed to the Boundary Enforcement assessment for GPT-4o in General scenarios. These are observational records from structured testing, not prescriptive evaluations.