# Risk-Adjusted Harm Scoring for Automated Red Teaming for LLMs in Financial Services

arXiv: https://arxiv.org/abs/2603.10807 · HTML: https://mlsanity.com/paper/2603.10807/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Safety alignment, refusal and guardrails, one of 12 directions in LLM safety and security.
- Builds on 21 of its 40 references that are in the map.
- First posted on arXiv 11 Mar 2026.

## Placed in

- LLM safety and security / [Safety alignment, refusal and guardrails](https://mlsanity.com/safety/safety-alignment-refusal-and-guardrails/)

## Builds on (most cited first, 21 of 21)

- [Universal and Transferable Adversarial Attacks on Aligned Language Models](https://mlsanity.com/paper/2307.15043/) (2023)
- [HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal](https://mlsanity.com/paper/2402.04249/) (2024)
- [Jailbreaking Black Box Large Language Models in Twenty Queries](https://mlsanity.com/paper/2310.08419/) (2023)
- [Red Teaming Language Models with Language Models](https://mlsanity.com/paper/2202.03286/) (2022)
- [DecodingTrust: A Comprehensive Assessment of Trustworthiness in GPT Models](https://mlsanity.com/paper/2306.11698/) (2023)
- [Great, Now Write an Article About That: The Crescendo Multi-Turn LLM Jailbreak Attack](https://mlsanity.com/paper/2404.01833/) (2024)
- [Large Language Models in Finance: A Survey](https://mlsanity.com/paper/2311.10723/) (2023)
- [WildTeaming at Scale: From In-the-Wild Jailbreaks to (Adversarially) Safer Language Models](https://mlsanity.com/paper/2406.18510/) (2024)
- [X-Teaming: Multi-Turn Jailbreaks and Defenses with Adaptive Multi-Agents](https://mlsanity.com/paper/2504.13203/) (2025)
- [BetterBench: Assessing AI Benchmarks, Uncovering Issues, and Establishing Best Practices](https://mlsanity.com/paper/2411.12990/) (2024)
- [LLM360: Towards Fully Transparent Open-Source LLMs](https://mlsanity.com/paper/2312.06550/) (2023)
- [A Survey of Large Language Models in Finance (FinLLMs)](https://mlsanity.com/paper/2402.02315/) (2024)
- [SG-Bench: Evaluating LLM Safety Generalization Across Diverse Tasks and Prompt Types](https://mlsanity.com/paper/2410.21965/) (2024)
- [When LLM Meets DRL: Advancing Jailbreaking Efficiency via DRL-guided Search](https://mlsanity.com/paper/2406.08705/) (2024)
- [Automated Red Teaming with GOAT: the Generative Offensive Agent Tester](https://mlsanity.com/paper/2410.01606/) (2024)
- [RedAgent: Red Teaming Large Language Models with Context-aware Autonomous Language Agent](https://mlsanity.com/paper/2407.16667/) (2024)
- [h4rm3l: A language for Composable Jailbreak Attack Synthesis](https://mlsanity.com/paper/2408.04811/) (2024)
- [BELLS: A Framework Towards Future Proof Benchmarks for the Evaluation of LLM Safeguards](https://mlsanity.com/paper/2406.01364/) (2024)
- [JADES: A Universal Framework for Jailbreak Assessment via Decompositional Scoring](https://mlsanity.com/paper/2508.20848/) (2025)
- [LeakAgent: RL-based Red-teaming Agent for LLM Privacy Leakage](https://mlsanity.com/paper/2412.05734/) (2024)
- [Learning to Conceal Risk: Controllable Multi-turn Red Teaming for LLMs in the Financial Domain](https://mlsanity.com/paper/2509.10546/) (2025)

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

- [FinRT: Distilling Adaptive Red-Teaming Strategies into Reusable Adversarial Generators in Consumer Finance](https://mlsanity.com/paper/2609.36474/) (2026)

## Abstract (arXiv)

Existing LLM safety evaluations rely on binary attack-success rates and domain-agnostic taxonomies, leaving regulated Banking, Financial Services, and Insurance (BFSI) deployments exposed to failures elicited through legally or professionally plausible framing. We introduce RAHS (Risk-Adjusted Harm Score), a risk-sensitive metric jointly capturing disclosure severity, disclaimer mitigation, and inter-judge agreement, and FinRedTeamBench, a 989-prompt benchmark spanning seven BFSI risk areas and 34 sub-categories mapped to regulatory frameworks. Evaluation uses an ensemble of three heterogeneous LLM judges, validated against human experts, and an adaptive multi-turn red-teaming pipeline. On nine open-weight models, RAHS preserves separation under near-ceiling ASR, ranking is stable under hyperparameter sweeps, and multi-turn pressure drives not only more jailbreaks but more operationally severe disclosures, exposing failure modes that single-turn, domain-agnostic evaluations cannot reveal.