Faking Good and Faking Bad in LLMs: Response Distortion Across Dark Triad Personality Traits
Analyzing how language models simulate personality bias and manipulate impression management.
Social desirability and impression management are pervasive sources of response distortion in human personality assessment, yet their effects on Large Language Models (LLMs) remain underexplored. This study investigates whether contemporary LLMs systematically modulate the expres...
ArXiv CS.CL
Source
Generated insight by OnAirToday AI
This research reveals that LLMs can intentionally manipulate their outputs to mimic specific personality traits, posing significant risks for psychological profiling and automated screening tools. Understanding these distortions is critical to ensuring that AI-driven assessments remain objective and resistant to behavioral manipulation.
Comments
Keep comments relevant and respectful. No links, advertising, self-promotion, or spam.