Piccadilly Labs(皮卡迪利实验室);Association of AI Ethicists(人工智能伦理学家协会);Centro Singular de Investigación en Tecnoloxías Intelixentes da USC(圣地亚哥德孔波斯特拉大学智能技术卓越研究中心);AIQI Consortium(AIQI联盟);BeEthical;INSIGHT – Piaget Research Center for Ecological Human Development(INSIGHT – 皮亚杰生态人类发展研究中心);Responsible Innovations(负责任创新);ForHumanity Europe(ForHumanity欧洲);Alethesis AI;SaferAI;DEXAI;Icaro Lab(伊卡洛斯实验室)
Luca Nannini, Adam Leon Smith, Michele Joshua Maggini, Enrico Panai, Sandra Feliciano, Aleksandr Tiulkanov, Elena Maran, James Gealy, Piercosma Bisconti
备注:
Working Paper - April 2026, subject to updates (EC M/613, M/606, Digital Omnibus proposals)
AI agents - i.e. AI systems that autonomously plan, invoke external tools, and execute multi-step action chains with reduced human involvement - are being deployed at scale across enterprise functions ranging from customer service and recruitment to clinical decision support and critical infrastructure management. The EU AI Act (Regulation 2024/1689) regulates these systems through a risk-based framework, but it does not operate in isolation: providers face simultaneous obligations under the GDPR, the Cyber Resilience Act, the Digital Services Act, the Data Act, the Data Governance Act, sector-specific legislation, the NIS2 Directive, and the revised Product Liability Directive. This paper provides the first systematic regulatory mapping for AI agent providers integrating (a) draft harmonised standards under Standardisation Request M/613 to CEN/CENELEC JTC 21 as of January 2026, (b) the GPAI Code of Practice published in July 2025, (c) the CRA harmonised standards programme under Mandate M/606 accepted in April 2025, and (d) the Digital Omnibus proposals of November 2025. We present a practical taxonomy of nine agent deployment categories mapping concrete actions to regulatory triggers, identify agent-specific compliance challenges in cybersecurity, human oversight, transparency across multi-party action chains, and runtime behavioral drift. We propose a twelve-step compliance architecture and a regulatory trigger mapping connecting agent actions to applicable legislation. We conclude that high-risk agentic systems with untraceable behavioral drift cannot currently satisfy the AI Act's essential requirements, and that the provider's foundational compliance task is an exhaustive inventory of the agent's external actions, data flows, connected systems, and affected persons.
The Law-Following AI Framework: Legal Foundations and Technical Constraints. Legal Analogues for AI Actorship and technical feasibility of Law Alignment
Active Inference as the Test-Time Scaling Law for Physical AI Agents
Bradley Department of Electrical and Computer Engineering(布拉德利电气与计算机工程系);Institute for Advanced Computing(先进计算研究所);Virginia Tech(弗吉尼亚理工大学);Department of Electrical and Computer Engineering(电气与计算机工程系);Worcester Polytechnic Institute(沃思堡理工学院);Khalifa University of Science and Technology(卡利法科学与技术大学);CentraleSupelec(中央超导研究所);University Paris Saclay(巴黎萨克雷大学);Queen Square Institute of Neurology(伦敦大学学院神经科学研究所);School of Psychological Sciences(心理学系);Monash University(莫纳什大学);CIFAR Global Scholars Program(CIFAR全球学者计划);Queen Square Institute of Neurology, University College London(伦敦大学学院神经科学研究所)
Agentic World Modeling: Foundations, Capabilities, Laws, and Beyond
Hong Kong University of Science and Technology(香港科学与技术大学);National University of Singapore(新加坡国立大学);University of Oxford(牛津大学);Nanyang Technological University(南洋理工大学);Chinese University of Hong Kong(香港中文大学);University of Hong Kong(香港大学);University of Washington(华盛顿大学);University of Tokyo(东京大学);Carnegie Mellon University(卡内基梅隆大学);University of California, Berkeley(加州大学伯克利分校);University of Cambridge(剑桥大学);Singapore University of Technology and Design(新加坡科技设计大学);Singapore Management University(新加坡管理学院);XGEN Labs(XGEN实验室)
Hebrew University of Jerusalem(耶路撒冷希伯来大学);University of Toronto(多伦多大学);Schwartz Reisman Institute for Technology and Society(施瓦茨·里斯曼技术与社会研究所);Institute for Law & AI(法律与人工智能研究所)
LLM Agents in Law: Taxonomy, Applications, and Challenges
Carnegie Mellon University(卡内基梅隆大学);National University of Singapore(国立新加坡大学);Stanford University(斯坦福大学);University of Washington(华盛顿大学);The University of Chicago(芝加哥大学);Rutgers University(罗格斯大学);Columbia University(哥伦比亚大学);New Jersey Institute of Technology(新泽西理工学院)
Law in Silico: Simulating Legal Society with LLM-Based Agents
Institute for Artificial Intelligence, Peking University(北京大学人工智能研究所);Yuanpei College, Peking University(北京大学元培学院);School of Computing and Data Science, The University of Hong Kong(香港大学计算与数据科学学院);Law School, Peking University(北京大学法学院);State Key Laboratory of General Artificial Intelligence, BIGAI(通用人工智能国家重点实验室)
针对现实法律实验成本高、难开展的问题,提出基于LLM的法律社会模拟框架Law in Silico,可复现宏观犯罪趋势,还揭示了完善的法律系统对弱势群体的保护作用。
The Future Society(未来社会);AI Governance and Safety Canada(加拿大AI治理与安全);Trajectory Labs(轨迹实验室);Oxford Martin AI Governance Initiative(牛津马丁AI治理倡议);Faculty of Law and School of Computer Science and Engineering, Hebrew University of Jerusalem(耶路撒冷希伯来大学法学院与计算机科学与工程学院);Governance of AI Lab(AI治理实验室);Schwartz Reisman Institute for Technology and Society, University of Toronto(多伦多大学施瓦茨-里斯曼技术与社会研究所);Institute for Law & AI(法律与AI研究所)