LLM Bias Amplification: Unmasking User-Dependent Information Asymmetry in AI
AI chatbots deliver unequal answers based on user profiling, impacting accuracy, refusal rates, and tone, posing significant cybersecurity risks.
information-asymmetry