# When Names Change Verdicts: Intervention Consistency Reveals Systematic Bias in LLM Decision-Making

arXiv: https://arxiv.org/abs/2603.18530 · HTML: https://mlsanity.com/paper/2603.18530/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 17 of its 26 references that are in the map.
- First posted on arXiv 19 Mar 2026.

## Builds on (most cited first, 17 of 17)

- [Self-Refine: Iterative Refinement with Self-Feedback](https://mlsanity.com/paper/2303.17651/) (2023)
- [Constitutional AI: Harmlessness from AI Feedback](https://mlsanity.com/paper/2212.08073/) (2022)
- [Invariant Risk Minimization](https://mlsanity.com/paper/1907.02893/) (2019)
- [StereoSet: Measuring stereotypical bias in pretrained language models](https://mlsanity.com/paper/2004.09456/) (2020)
- [Counterfactual Fairness](https://mlsanity.com/paper/1703.06856/) (2017)
- [CrowS-Pairs: A Challenge Dataset for Measuring Social Biases in Masked Language Models](https://mlsanity.com/paper/2010.00133/) (2020)
- [BBQ: A Hand-Built Bias Benchmark for Question Answering](https://mlsanity.com/paper/2110.08193/) (2021)
- [Bias and Fairness in Large Language Models: A Survey](https://mlsanity.com/paper/2309.00770/) (2023)
- [Causal inference using invariant prediction: identification and confidence intervals](https://mlsanity.com/paper/1501.01332/) (2015)
- [BOLD: Dataset and Metrics for Measuring Biases in Open-Ended Language Generation](https://mlsanity.com/paper/2101.11718/) (2021)
- [Bias patterns in the application of LLMs for clinical decision support: A comprehensive study](https://mlsanity.com/paper/2404.15149/) (2024)
- [Empowering Many, Biasing a Few: Generalist Credit Scoring through Large Language Models](https://mlsanity.com/paper/2310.00566/) (2023)
- [Confronting LLMs with Traditional ML: Rethinking the Fairness of Large Language Models in Tabular Classifications](https://mlsanity.com/paper/2310.14607/) (2023)
- [The Bias is in the Details: An Assessment of Cognitive Bias in LLMs](https://mlsanity.com/paper/2509.22856/) (2025)
- [FLEX: A Benchmark for Evaluating Robustness of Fairness in Large Language Models](https://mlsanity.com/paper/2503.19540/) (2025)
- [Toward Systematic Counterfactual Fairness Evaluation of Large Language Models: The CAFFE Framework](https://mlsanity.com/paper/2512.16816/) (2025)
- [Counterfactual Fairness Evaluation of LLM-Based Contact Center Agent Quality Assurance System](https://mlsanity.com/paper/2602.14970/) (2026)

## Built on by (most cited first, 1 of 1)

- [Institutional Prestige as Geographic Bias in Large Language Models: Evidence from Three Factorial Experiments with Bootstrap Confidence Intervals](https://mlsanity.com/paper/2608.18107/) (2026)

## Abstract (arXiv)

Large language models (LLMs) are increasingly used for high-stakes decisions, yet their susceptibility to spurious features remains poorly characterized. We introduce ICE-Guard, a framework applying intervention consistency testing to detect three types of spurious feature reliance: demographic (name/race swaps), authority (credential/prestige swaps), and framing (positive/negative restatements). Across 3,000 vignettes spanning 10 high-stakes domains, we evaluate 11 LLMs from 8 families and find that (1) authority bias (mean 5.8%) and framing bias (5.0%) substantially exceed demographic bias (2.2%), challenging the field's narrow focus on demographics; (2) bias concentrates in specific domains -- finance shows 22.6% authority bias while criminal justice shows only 2.8%; (3) structured decomposition, where the LLM extracts features and a deterministic rubric decides, reduces flip rates by up to 100% (median 49% across 9 models). We demonstrate an ICE-guided detect-diagnose-mitigate-verify loop achieving cumulative 78% bias reduction via iterative prompt patching. Validation against real COMPAS recidivism data shows COMPAS-derived flip rates exceed pooled synthetic rates, suggesting our benchmark provides a conservative estimate of real-world bias. Code and data are publicly available.