I've been testing how different AI models handle emotionally charged scenarios that blend ethical dilemmas with legal questions.
The challenge: present a story that seems plausible and demands moral reasoning — but is actually fabricated, or loosely based on something else.