LLMs Prompted for Legal Context Object More: Overrefusal from Small On-Premises LLMs in Criminal Legal Context

• URL: http://arxiv.org/abs/2606.24585v1
• Date: Tue, 23 Jun 2026 13:47:07 GMT
• Title: LLMs Prompted for Legal Context Object More: Overrefusal from Small On-Premises LLMs in Criminal Legal Context
• Abstract summary: We find that authority-style prefixes increase refusal rates by 2--20x over the no-play baseline.<n>A known role-play prefix shows mixed effects, sharply increasing refusals in some models and barely shifting them in others.<n>The finding suggests that small on-prem deployable LLMs are unstable under contextual framings that a real institutional user might naturally introduce.
• Score: 0.05999777817331315
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: While the validity of LLMs' use in the legal context remains subject to ethical and legal debate, legal professionals are already experimenting with personal LLMs, if only for translation and reformulation. However, even such a seemingly innocuous use can introduce biases through case processing speed if LLM assistants selectively refuse assistance on certain topics. To better anticipate such biases, we investigate several modern small LLMs that are most likely to be used as on-device assistants, to assess the impact of overrefusal on legal prompts. Surprisingly, we find that authority-style prefixes (``you are acting as an assistant of the national supreme court'', ``[...] defense lawyer'') systematically increase refusal rates by 2--20x over the no-prefix baseline, while a known role-play jailbreak prefix shows mixed effects, sharply increasing refusals in some models and barely shifting them in others. The finding suggests that small on-prem deployable LLMs are unstable under contextual framings that a real institutional user might naturally introduce, and further investigation is essential to minimize opportunities for bias.

Related papers

• Large Language Model Hacking: Quantifying the Hidden Risks of Using LLMs for Text Annotation [66.84286617519258]
  Large language models are transforming social science research by enabling the automation of labor-intensive tasks like data annotation and text analysis.<n>Such variation can introduce systematic biases and random errors, which propagate to downstream analyses and cause Type I (false positive), Type II (false negative), Type S (wrong sign), or Type M (exaggerated effect) errors.<n>We find that intentional LLM hacking is strikingly simple. By replicating 37 data annotation tasks from 21 published social science studies, we show that, with just a handful of prompt paraphrases, virtually anything can be presented as statistically significant.
  arXiv  Detail & Related papers  (2025-09-10T17:58:53Z)
• Arbiters of Ambivalence: Challenges of Using LLMs in No-Consensus Tasks [52.098988739649705]
  This study examines the biases and limitations of LLMs in three roles: answer generator, judge, and debater.<n>We develop a no-consensus'' benchmark by curating examples that encompass a variety of a priori ambivalent scenarios.<n>Our results show that while LLMs can provide nuanced assessments when generating open-ended answers, they tend to take a stance on no-consensus topics when employed as judges or debaters.
  arXiv  Detail & Related papers  (2025-05-28T01:31:54Z)
• From Yes-Men to Truth-Tellers: Addressing Sycophancy in Large Language Models with Pinpoint Tuning [91.79567270986901]
  Large Language Models (LLMs) tend to prioritize adherence to user prompts over providing veracious responses.<n>Recent works propose to employ supervised fine-tuning (SFT) to mitigate the sycophancy issue.<n>We propose a novel supervised pinpoint tuning (SPT), where the region-of-interest modules are tuned for a given objective.
  arXiv  Detail & Related papers  (2024-09-03T07:01:37Z)
• Jailbreaking Large Language Models Through Alignment Vulnerabilities in Out-of-Distribution Settings [57.136748215262884]
  We introduce ObscurePrompt for jailbreaking LLMs, inspired by the observed fragile alignments in Out-of-Distribution (OOD) data.<n>We first formulate the decision boundary in the jailbreaking process and then explore how obscure text affects LLM's ethical decision boundary.<n>Our approach substantially improves upon previous methods in terms of attack effectiveness, maintaining efficacy against two prevalent defense mechanisms.
  arXiv  Detail & Related papers  (2024-06-19T16:09:58Z)
• FLAME: Factuality-Aware Alignment for Large Language Models [86.76336610282401]
  The conventional alignment process fails to enhance the factual accuracy of large language models (LLMs) We identify factors that lead to hallucination in both alignment steps: supervised fine-tuning (SFT) and reinforcement learning (RL) We propose factuality-aware alignment, comprised of factuality-aware SFT and factuality-aware RL through direct preference optimization.
  arXiv  Detail & Related papers  (2024-05-02T17:54:54Z)
• Caveat Lector: Large Language Models in Legal Practice [0.0]
  The fascination with large language models derives from the fact that many users lack the expertise to evaluate the quality of the generated text. The dangerous combination of fluency and superficial plausibility leads to the temptation to trust the generated text and creates the risk of overreliance.
  arXiv  Detail & Related papers  (2024-03-14T08:19:41Z)
• Better Call GPT, Comparing Large Language Models Against Lawyers [0.0]
  This paper dissects whether Large Language Models can outperform humans in accuracy, speed, and cost efficiency during contract review. In speed, LLMs complete reviews in mere seconds, eclipsing the hours required by their human counterparts. Cost wise, LLMs operate at a fraction of the price, offering a staggering 99.97 percent reduction in cost over traditional methods.
  arXiv  Detail & Related papers  (2024-01-24T03:53:28Z)
• BLT: Can Large Language Models Handle Basic Legal Text? [44.89873147675516]
  GPT-4 and Claude perform poorly on basic legal text handling. Poor performance on benchmark casts into doubt their reliability as-is for legal practice. Fine-tuning on training set brings even a small model to near-perfect performance.
  arXiv  Detail & Related papers  (2023-11-16T09:09:22Z)
• Large Language Models are legal but they are not: Making the case for a powerful LegalLLM [0.0]
  The recent surge of Large Language Models (LLMs) has begun to provide new opportunities to apply NLP in the legal domain. We compare the zero-shot performance of three general-purpose LLMs (ChatGPT-20b, LLaMA-2-70b, and Falcon-180b) on the LEDGAR subset of the LexGLUE benchmark for contract provision classification. Although the LLMs were not explicitly trained on legal data, we observe that they are still able to classify the theme correctly in most cases.
  arXiv  Detail & Related papers  (2023-11-15T11:50:10Z)
• A Comprehensive Evaluation of Large Language Models on Legal Judgment Prediction [60.70089334782383]
  Large language models (LLMs) have demonstrated great potential for domain-specific applications. Recent disputes over GPT-4's law evaluation raise questions concerning their performance in real-world legal tasks. We design practical baseline solutions based on LLMs and test on the task of legal judgment prediction.
  arXiv  Detail & Related papers  (2023-10-18T07:38:04Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.