📄 中文摘要
一项实验显示，GPT-5 在法律推理方面的表现显著优于联邦法官，准确率达到 100%，而法官的准确率仅为 52%。该实验旨在评估人工智能在法律领域的应用潜力，尤其是在分析复杂法律问题和提供法律建议方面。研究人员认为，GPT-5 的成功表明，人工智能可以在法律行业中发挥重要作用，可能会改变法律实践的未来。此结果引发了对 AI 在法律决策中应用的广泛讨论，尤其是在公正性和透明度方面的潜在影响。
GPT-5 在法律推理实验中以 100% 的准确率超越联邦法官的 52%
📄 English Summary
GPT-5 outperforms federal judges 100% to 52% in legal reasoning experiment
An experiment revealed that GPT-5 significantly outperformed federal judges in legal reasoning, achieving an accuracy rate of 100% compared to the judges' 52%. The study aimed to assess the potential applications of artificial intelligence in the legal field, particularly in analyzing complex legal issues and providing legal advice. Researchers believe that GPT-5's success indicates that AI can play a crucial role in the legal industry, potentially transforming the future of legal practice. This outcome has sparked widespread discussions about the implications of AI in legal decision-making, especially concerning fairness and transparency.