{
  "id": 13482,
  "url": "https://arxiv.org/abs/2607.21063v1",
  "title": "QuantiBias: Benchmarking Quantization-Induced Bias in LLMs",
  "summary": "Almost every large language model that reaches a broad audience is quantized: trained in full precision, then compressed for efficiency. This step is assumed harmless and its safety is rarely re-checked. We find its principal side effect is increased bias that standard safety evaluation misses. Holding the model, its training, and the prompts fixed, a quantized model still refuses harmful requests, still avoids over-refusing benign prompts, and still selects the unbiased multiple-choice answer.",
  "authors": "Emilio Ferrara",
  "category": "research",
  "topics": "bias-fairness",
  "orgs": null,
  "regions": null,
  "published_at": "2026-07-23T08:56:11.000Z",
  "fetched_at": "2026-07-25T05:10:48.796Z",
  "source_slug": "x-arxiv-cs-hc",
  "source_name": "arXiv cs.HC",
  "source_homepage": "https://arxiv.org/list/cs.HC/recent",
  "ethics_ai_record_url": "https://ethics.ai/record/13482",
  "original_url": "https://arxiv.org/abs/2607.21063v1",
  "evidence_status": "source-only",
  "attribution": "via ethics.ai"
}