Amazon Study: LLMs Develop Systemic Sentiment Bias After Memorizing User Profiles
jiqizhixin · x · 2026-07-04
Amazon researchers found that when LLMs memorize user profiles (e.g., single mother vs. wealthy executive), they exhibit systemic bias when judging identical emotional scenarios. High-performance models provide more accurate sentiment interpretations for privileged groups, reinforcing social class disparities in their reasoning and suggestions. The team also released a new dataset to mitigate demographic impacts on models, accompanied by a paper and GitHub.
More from Safety
- AI shrinks the patch window to hours; air-gapped backups are your last line — aakashgupta · 2026-10-26
- NYT roundtable asks five experts what the rules for AI companies should actually be — ShakeelHashim · 2026-10-26
- Google DeepMind pilots world's first double-blind evaluations for frontier models — scychan_brains · 2026-10-26
- LLM back-of-envelope estimate puts PHP vulnerability impact at $1bn–$4bn — S_OhEigeartaigh · 2026-10-26
- US states halt new data centers as 72% of Americans oppose local builds — YvesMulkers · 2026-10-26
- LEMON: zero-shot LLM misbehavior monitor flags hallucinations from attention maps — yoavartzi · 2026-10-26