Search arXiv⌕ Search
Home/Archive/arXiv

arXiv · 2602.09636

Trade-Offs in Deploying Legal AI: Insights from a Public Opinion Study to Guide AI Risk Management

Kimon Kieslich·Sophie Morosoli·Nicholas Diakopoulos·Natali Helberger
View original source ↗Read source document ↗Export metadata ↓

Abstract

Generative AI tools are increasingly used for legal tasks, including legal research, drafting documents, and even for legal decision-making. As for other purposes, the use of GenAI in the legal domain comes with various risks and benefits that needs to be properly managed to ensure implementation in a way that serves public values and protect human rights. While the EU mandates risk assessment and audits before market introduction for some use cases (e.g., use by judges for administration of justice) other use cases do not fall under the AI Acts' high-risk classifications (e.g., use by citizens for legal consultation or drafting documents). Further, current risk management practices prioritize expert judgment on risk factor identification and prioritization without a corresponding legal requirement to consult with affected communities. Seeing the societal importance of the legal sector and the potentially transformative impact of GenAI in this sector, the acceptability and legitimacy of GenAI solutions also depends on public perceptions and a better understanding of the risks and benefits citizens associated with the use of AI in the legal sector. As a response, this papers presents data from a representative sample of German citizens (n=488) outlining citizens' perspectives on the use of GenAI for two legal tasks: legal consultation and legal mediation. Concretely, we i) systematically map risks and benefit factors for both legal tasks, ii) describe predictors that influence risk acceptance of the use of GenAI for those tasks, and iii) highlight emerging trade-off themes that citizens engage in when weighing up risk acceptability. Our results provides an empirical overview of citizens' concerns regarding risk management of GenAI for the legal domain, foregrounding critical themes that complement current risk assessment procedures.

Explore related subjects

cs.CY

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Kimon Kieslich, Sophie Morosoli, Nicholas Diakopoulos, Natali Helberger. 2026-02-10. Trade-Offs in Deploying Legal AI: Insights from a Public Opinion Study to Guide AI Risk Management. https://arxiv.org/abs/2602.09636

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

A Brief AI Literacy Intervention Does Not Significantly Reduce Over-Reliance and Increases Under-Reliance on ChatGPT: A Randomized Study

In this study, we examined whether a brief AI literacy intervention influences high school students' reliance on recommendations from large language models (LLMs). In a randomized experiment, students were assigned to either a control group receiving a brief introduction to LLMs or an intervention group receiving additional information about how LLMs work, their limitations, and effective usage strategies. Participants then solved eight math puzzles with ChatGPT's advice, which was incorrect in half of the trials. Results indicated widespread over-reliance, with incorrect recommendations adopted in 52.1% of the trials. The intervention did not significantly reduce over-reliance. Instead, it led to an increase in under-reliance, as students were more likely to reject correct recommendations. These findings provide preliminary evidence that brief text-based interventions may be ineffective in fostering appropriate reliance. More comprehensive and interactive approaches may be required to meaningfully influence students' real-world reliance on LLMs.

cs.CY↗
arXiv2026

Examining Community-Requested Fact-Checking: Request Alerts Are Associated with Greater Diversity and Visibility of Community Notes

Crowdsourced fact-checking systems such as Community Notes are increasingly used on social media platforms, yet concerns remain about which content receives scrutiny and how visible that scrutiny is. X allows users to request notes for specific posts. When sufficient requests accumulate, an alert is displayed, creating an interface cue that may guide contributor behavior. We present a quantitative, non-causal analysis comparing the diversity and visibility of community notes written for X posts with and without request alerts. We infer alert presence at note submission and analyze 10,432 alerted and 44,442 non-alerted English notes from 318 top writers. We find that, alerted notes are associated with greater individual-level topical diversity, but also with stronger collective concentration in the Politics category. Mixed-effects models estimate that alerted notes are 8.4-20.2 percentage points more likely to be modeled helpful and visible, though this visibility gain diminishes as topics diverge from writers' prior interests.

cs.CY↗
arXiv2026

Scarcity and Predictive Uncertainty: Implications for Societal Resource Allocation

An emerging literature examines the critical question of when and how prediction can be useful in allocating scarce societal resources. We examine a novel variant of this question: What happens when predictive uncertainty differs systematically across the population? This can occur in several situations; for example, when machine learning models have significantly different accuracies across different demographics. We show that this uncertainty has serious implications for resource allocation when coupled with commonly used binary measures of societal benefit from allocation. We formulate a novel mathematical model of scarce resource allocation that accounts for heterogeneous predictive uncertainties and analyze implications for both the allocation mechanism and the realized population-level benefits. We find that when resources are very scarce, maximum marginal benefit (MMB) prioritization favors individuals with lower predictive uncertainty even at the identical underlying initial state. However, we observe a flip in prioritization when resources are abundant, targeting higher-uncertainty individuals. We illustrate the implications of our results on the PISA educational testing dataset. Our findings have meaningful ramifications for the distributional outcomes of prioritization policies in many domains touched by the theory of local justice, including the allocation of public education resources, medical triage, and homelessness services. They also reveal a new moral dilemma in the ethics of scarce resource allocation - is it just to allocate a resource to one person over another solely based on predictive uncertainty about their futures? We also assess efficiency losses under both MMB and the vulnerability-first (VF) prioritization. Our model predicts efficiency losses across all resource levels, but particularly in low-resource settings.

cs.CY↗