SOURCE-LINKED INTELLIGENCE

Harm Laundering in GPT Models: Evidence That Gender Discrimination Is Transformed Rather Than Reduced Across Safety-Trained Generations

arXiv · AI, language, vision and robotics · article · Sep 17, 2026 · UTC

Safety evaluations for large language models rely on surface-form classifiers that report declining harm scores across model generations. We provide evidence that this methodology is systematically incomplete: explicit discriminatory content is transformed rather than removed. We call this \emph{harm laundering}. Analysing 450,000 gender-directed completions across 15 models spanning GPT-2 through to GPT-5 (OpenAI GPT lineage; three demographic conditions), we show that sexual violence clusters prevalent in GPT-2 women-directed output disappear by GPT-4, while men-directed completions gain pos

Read original source ↗ Open in workspace

recordType

paper

region

Global