Generative AI (GenAI) tools are transforming critical societal domains, including the legal sector. While these tools create opportunities such as increased efficiency and potential improvements in access to justice, they also present new challenges, such as the risk of inaccurate legal advice and questions about the legitimacy of legal decisions. However, the full impact remains to be seen and ultimately depends on the way GenAI tools are implemented and used by both, legal professionals and citizens. This makes anticipating and managing the positive and negative impacts of GenAI use in the legal domain challenging but also important to guide the digital transformation of the legal sector into a societally desirable direction. In this paper, we set out to explore the spectrum of possible impacts of GenAI in the legal domain, examining how this technology is anticipated being used and the potential implications this might have for the legal sector and society. Using a scenario writing method, we surveyed participants in the EU and US including both citizens and legal professionals about the potential impact of generative AI on legal conflict resolution. Respondents were tasked with writing a narrative drawing on their experience or expertise about a future in which AI is used throughout the legal process. We qualitatively analysed the prevalence of risk and benefit themes, as well as the types of anticipated legal tasks. We then compared these findings based on expertise status (legal experts versus citizens) and regional regulatory background (the EU with the EU AI Act versus the US with an industry self-regulatory approach). Finally, we describe the emerging trade-offs that will affect decision-makers in the legal sector.

翻译：生成式人工智能（GenAI）工具正在改变包括法律领域在内的关键社会领域。这些工具在创造机遇（如提升效率、改善司法可及性）的同时，也带来了新的挑战，例如法律建议不准确的风险以及法律决策合法性的质疑。然而，其全面影响仍有待观察，最终取决于法律专业人士和公民对GenAI工具的实施与使用方式。这使得预测和管理GenAI在法律领域应用的积极与消极影响具有挑战性，但对于引导法律行业数字化转型朝着社会期望的方向发展至关重要。本文旨在探索GenAI在法律领域可能产生的影响谱系，考察该技术的预期使用方式及其对法律行业和社会的潜在影响。通过场景写作方法，我们在欧盟和美国调查了包括公民和法律专业人士在内的参与者，探讨生成式AI对法律冲突解决的潜在影响。受访者需根据自身经验或专业知识，撰写关于AI全程参与法律程序的未来叙事。我们对风险与收益主题的普遍性以及预期法律任务的类型进行了定性分析，并依据专业身份（法律专家与公民）和区域监管背景（欧盟的《人工智能法案》与美国的行业自律模式）对这些发现进行了比较。最后，我们阐述了将影响法律领域决策者的新兴权衡关系。