From Trolley Problem to Robot Sociopaths: Empathy's Role in AI Safety

Scobleizer · x · 2026-08-10

USC PhD researcher Roshni Lulla explores the implications of human empathy mechanisms for AI safety. She divides empathy into cognitive empathy (understanding thoughts) and affective empathy (sharing feelings).

She notes that a lack of affective empathy often leads to antisocial behavior in humans. If AI systems possess powerful capabilities without genuine empathy mechanisms, they might make dangerous, 'robot sociopath' decisions in moral dilemmas. Thus, applying neuroscience and behavioral research to AI alignment is crucial for building safe and ethical systems.

Original post →

More from AGI Musings

AGI Musings channel →