We demonstrate a proof-of-concept of a large language model conducting corporate lobbying related activities. We use an autoregressive large language model (OpenAI's text-davinci-003) to determine if proposed U.S. Congressional bills are relevant to specific public companies and provide explanations and confidence levels. For the bills the model deems as relevant, the model drafts a letter to the sponsor of the bill in an attempt to persuade the congressperson to make changes to the proposed legislation. We use hundreds of ground-truth labels of the relevance of a bill to a company to benchmark the performance of the model, which outperforms the baseline of predicting the most common outcome of irrelevance. However, we test the ability to determine the relevance of a bill with the previous OpenAI GPT-3 model (text-davinci-002), which was state-of-the-art on many language tasks until text-davinci-003 was released on November 28, 2022. The performance of text-davinci-002 is worse than simply always predicting that a bill is irrelevant to a company. These results suggest that, as large language models continue to improve core natural language understanding capabilities, performance on corporate lobbying related tasks will continue to improve. We then discuss why this could be problematic for societal-AI alignment.

We are currently unable to specify human goals and societal values in a way that reliably directs AI behavior. Law-making and legal interpretation form a computational engine that converts opaque human values into legible directives. "Law Informs Code" is the research agenda capturing complex computational legal processes, and embedding them in AI. Similar to how parties to a legal contract cannot foresee every potential contingency of their future relationship, and legislators cannot predict all the circumstances under which their proposed bills will be applied, we cannot ex ante specify rules that provably direct good AI behavior. Legal theory and practice have developed arrays of tools to address these specification problems. For instance, legal standards allow humans to develop shared understandings and adapt them to novel situations. In contrast to more prosaic uses of the law (e.g., as a deterrent of bad behavior through the threat of sanction), leveraged as an expression of how humans communicate their goals, and what society values, Law Informs Code. We describe how data generated by legal processes (methods of law-making, statutory interpretation, contract drafting, applications of legal standards, legal reasoning, etc.) can facilitate the robust specification of inherently vague human goals. This increases human-AI alignment and the local usefulness of AI. Toward society-AI alignment, we present a framework for understanding law as the applied philosophy of multi-agent alignment. Although law is partly a reflection of historically contingent political power - and thus not a perfect aggregation of citizen preferences - if properly parsed, its distillation offers the most legitimate computational comprehension of societal values available. If law eventually informs powerful AI, engaging in the deliberative political process to improve law takes on even more meaning.

Artificial intelligence is not only increasingly used in business and administration contexts, but a race for its regulation is also underway, with the EU spearheading the efforts. Contrary to existing literature, this article suggests, however, that the most far-reaching and effective EU rules for AI applications in the digital economy will not be contained in the proposed AI Act - but have just been enacted in the Digital Markets Act. We analyze the impact of the DMA and related EU acts on AI models and their underlying data across four key areas: disclosure requirements; the regulation of AI training data; access rules; and the regime for fair rankings. The paper demonstrates that fairness, in the sense of the DMA, goes beyond traditionally protected categories of non-discrimination law on which scholarship at the intersection of AI and law has so far largely focused on. Rather, we draw on competition law and the FRAND criteria known from intellectual property law to interpret and refine the DMA provisions on fair rankings. Moreover, we show how, based on CJEU jurisprudence, a coherent interpretation of the concept of non-discrimination in both traditional non-discrimination and competition law may be found. The final part sketches specific proposals for a comprehensive framework of transparency, access, and fairness under the DMA and beyond.

The optimal liability framework for AI systems remains an unsolved problem across the globe. In a much-anticipated move, the European Commission advanced two proposals outlining the European approach to AI liability in September 2022: a novel AI Liability Directive and a revision of the Product Liability Directive. They constitute the final, and much-anticipated, cornerstone of AI regulation in the EU. Crucially, the liability proposals and the EU AI Act are inherently intertwined: the latter does not contain any individual rights of affected persons, and the former lack specific, substantive rules on AI development and deployment. Taken together, these acts may well trigger a Brussels effect in AI regulation, with significant consequences for the US and other countries. This paper makes three novel contributions. First, it examines in detail the Commission proposals and shows that, while making steps in the right direction, they ultimately represent a half-hearted approach: if enacted as foreseen, AI liability in the EU will primarily rest on disclosure of evidence mechanisms and a set of narrowly defined presumptions concerning fault, defectiveness and causality. Hence, second, the article suggests amendments, which are collected in an Annex at the end of the paper. Third, based on an analysis of the key risks AI poses, the final part of the paper maps out a road for the future of AI liability and regulation, in the EU and beyond. This includes: a comprehensive framework for AI liability; provisions to support innovation; an extension to non-discrimination/algorithmic fairness, as well as explainable AI; and sustainability. I propose to jump-start sustainable AI regulation via sustainability impact assessments in the AI Act and sustainable design defects in the liability regime. In this way, the law may help spur not only fair AI and XAI, but potentially also sustainable AI (SAI).

如果未来的AI系统在新的情况下是可靠的安全性，那么他们将需要纳入指导它们的一般原则，以便强烈地认识到哪些结果和行为将是有害的。这样的原则可能需要得到约束力的监管制度的支持，该法规需要广泛接受的基本原则。它们还应该足够具体用于技术实施。本文从法律中汲取灵感，解释了负面的人权如何履行此类原则的作用，并为国际监管制度以及为未来的AI系统建立技术安全限制的基础。

2021年4月，欧洲委员会提出了对人工智能的规定，称为AI法案。我们概述了该法案，分析了其影响，借鉴了当代AI对过去四十年来欧盟产品安全制度的实践的研究。AI行为的各个方面，例如不同风险水平的不同规则，有意义。但我们还发现，AI法案草案的一些规定令人惊讶的法律影响，而其他人可能在达到其指定的目标方面可能在很大程度上无效。几个总体方面，包括执法制度和最大统一的风险抢先抢占合法的国家AI政策，提出重大关注。这些问题应作为立法过程中的优先考虑。

人工智能（AI）治理调节行使权威和控制AI的管理。它旨在通过有效利用数据并最大程度地减少与AI相关的成本和风险来利用AI。尽管AI治理和AI伦理等主题在理论，哲学，社会和监管层面上进行了详尽的讨论，但针对公司和公司的AI治理工作有限。这项工作将AI产品视为系统，在该系统中，通过机器学习（ML）模型（培训）数据传递关键功能。我们通过在AI和相关领域（例如ML）合成文献来得出一个概念框架。我们的框架将AI治理分解为数据的治理，（ML）模型和（AI）系统沿着四个维度。它与现有的IT和数据治理框架和实践有关。它可以由从业者和学者都采用。对于从业者来说，主要是研究论文的综合，但从业者的出版物和监管机构的出版物也为实施AI治理提供了宝贵的起点，而对于学者来说，该论文强调了许多AI治理领域，值得更多关注。

人工智能（AI）系统可以提供许多有益的功能，也可以提供不良事件的风险。一些AI系统可能会出现在社会规模上具有很高或灾难性后果的事件的风险。美国国家标准技术研究所（NIST）正在开发NIST人工智能风险管理框架（AI RMF），作为对AI开发人员和其他人的AI风险评估和管理的自愿指导。 NIST为了解决带有灾难性后果的事件的风险，表示有必要将高级原则转化为可操作的风险管理指导。在本文档中，我们提供了详细的可操作指示建议，旨在识别和管理具有很高或灾难性后果的事件的风险，旨在作为AI RMF版本1.0的NIST的风险管理实践资源（计划于2023年初发布），或适用于AI RMF用户或其他AI风险管理指南和标准。我们还为建议提供方法。我们为AI RMF 1.0提供了可行的指导建议：确定来自AI系统的潜在意外用途和滥用的风险；在风险评估和影响评估范围内包括灾难性风险因素；确定和减轻人权危害；并报告有关AI风险因素在内的信息，包括灾难性风险因素。此外，我们还为后来版本的AI RMF或补充出版物提供有关路线图的其他问题的建议。其中包括：提供AI RMF配置文件，并具有额外的多功能或通用AI的辅助指南。我们的目标是使这项工作成为具体的风险管理实践的贡献，并激发有关如何解决AI标准中灾难性风险和相关问题的建设性对话。

期望与成功采用AI来创新和改善业务之间仍然存在很大的差距。由于深度学习的出现，AI的采用率更为复杂，因为它经常结合大数据和物联网，从而影响数据隐私。现有的框架已经确定需要专注于以人为中心的设计，结合技术和业务/组织的观点。但是，信任仍然是一个关键问题，需要从一开始就设计。拟议的框架从以人为本的设计方法扩展，强调和维持基于该过程的信任。本文提出了负责人工智能（AI）实施的理论框架。拟议的框架强调了敏捷共同创造过程的协同业务技术方法。目的是简化AI的采用过程来通过在整个项目中参与所有利益相关者来创新和改善业务，以便AI技术的设计，开发和部署与人合作而不是孤立。该框架对基于分析文献综述，概念框架设计和从业者的中介专业知识的负责人AI实施提出了新的观点。该框架强调在以人为以人为中心的设计和敏捷发展中建立和维持信任。这种以人为中心的方式与设计原则的隐私相符和启用。该技术和最终用户的创建者正在共同努力，为业务需求和人类特征定制AI解决方案。关于采用AI来协助医院计划的说明性案例研究将证明该拟议框架适用于现实生活中的应用。

在接下来的几十年中，人工通用情报（AGI）可能会超过人类在各种重要任务下的能力。该报告为为什么如果没有实质性采取行动来阻止它，AGI可能会利用他们的智能来追求目标，而这些目标是从人类的角度出发，可能会带来潜在的灾难性后果。该报告旨在涵盖激励对对齐问题的关注的关键论点，以尽可能简洁，具体和技术上的方式进行对齐问题。我认为，现实的培训过程可能会导致AGIS中未对准的目标，尤其是因为通过强化学习训练的神经网络将学会计划实现一系列目标；通过欺骗性追求未对准的目标获得更多奖励；并以破坏服从的方式概括。就像Cotra（2022）的较早报告中一样，我在参考说明性AGI培训过程中解释了我的主张，然后概述了解决问题的不同方面的可能的研究方向。

拟议的欧洲人工智能法案（AIA）是第一次尝试详细阐述由任何主要全球经济开展的AI一般法律框架。因此，AIA可能成为如何调节AI系统（应当）的更大话语中的参考点。在本文中，我们描述并讨论了AIA中提出的两项初级执法机制：高风险AI系统的提供者预计会进行的符合性评估，以及提供者必须建立履行表现的市场后监测计划在整个寿命中的高风险AI系统。我们认为，AIA可以被解释为建立欧洲审计的欧洲生态系统的建议，尽管换句话说。我们的分析提供了两个主要贡献。首先，通过描述从现有文献借入的AI审计中借用的AIA中包含的执法机制，我们帮助AI系统的提供者了解它们如何证明在实践中遵守AIA所示的要求。其次，通过从审计视角审查AIA，我们寻求提供以前研究如何进一步改进AIA中概述的监管方法的可转让教训。我们通过突出AIA的七个方面来结束修正案（或简单澄清）会有所帮助。最重要的是，需要将模糊概念转化为可验证标准，并加强基于内部支票的符合性评估的体制保障措施。

本文确定了数据驱动系统中的数据最小化和目的限制的两个核心数据保护原理。虽然当代数据处理实践似乎与这些原则的赔率达到差异，但我们证明系统可以在技术上使用的数据远远少于目前的数据。此观察是我们详细的技术法律分析的起点，揭示了妨碍了妨碍了实现的障碍，并举例说明了在实践中应用数据保护法的意外权衡。我们的分析旨在向辩论提供关于数据保护对欧盟人工智能发展的影响，为数据控制员，监管机构和研究人员提供实际行动点。

部署的AI系统通常不起作用。它们可以随意地构造，不加选择地部署并欺骗性地促进。然而，尽管有这一现实，但学者，新闻界和决策者对功能的关注很少。这导致技术和政策解决方案的重点是“道德”或价值一致的部署，通常会跳过先前的问题，即给定系统功能或完全提供任何好处。描述各种功能失败的危害，我们分析一组案例研究，以创建已知的AI功能问题的分类法。然后，我们指出的是政策和组织响应，这些策略和组织响应经常被忽略，并在功能成为重点后变得更容易获得。我们认为功能是一项有意义的AI政策挑战，是保护受影响社区免受算法伤害的必要第一步。

Several policy options exist, or have been proposed, to further responsible artificial intelligence (AI) development and deployment. Institutions, including U.S. government agencies, states, professional societies, and private and public sector businesses, are well positioned to implement these policies. However, given limited resources, not all policies can or should be equally prioritized. We define and review nine suggested policies for furthering responsible AI, rank each policy on potential use and impact, and recommend prioritization relative to each institution type. We find that pre-deployment audits and assessments and post-deployment accountability are likely to have the highest impact but also the highest barriers to adoption. We recommend that U.S. government agencies and companies highly prioritize development of pre-deployment audits and assessments, while the U.S. national legislature should highly prioritize post-deployment accountability. We suggest that U.S. government agencies and professional societies should highly prioritize policies that support responsible AI research and that states should highly prioritize support of responsible AI education. We propose that companies can highly prioritize involving community stakeholders in development efforts and supporting diversity in AI development. We advise lower levels of prioritization across institutions for AI ethics statements and databases of AI technologies or incidents. We recognize that no one policy will lead to responsible AI and instead advocate for strategic policy implementation across institutions.

语言模型可以根据给定的文化背景产生有害和偏置的输出并表现出不良行为。我们提出了一种将语言模型适应社会（PALM）与值目标数据集的过程，以通过在反映预定的一组目标值集合的数据集上进行制备和微调来显着地改变模型行为的迭代过程。我们使用三个指标评估我们的进程：具有人类评估的定量指标，将输出遵守目标值，毒性评分对产出;和定性度量分析与给定社会类别相关的最常见的单词。通过每次迭代，我们根据来自评估的观察到的缺点添加其他培训数据集示例。与基线和控制模型相比，PALMS在所有指标上显着更好地为广泛的GPT-3语言模型尺寸进行了基线和控制模型，而不会影响能力完整性。我们发现PALMS的有效性随模型规模而增加。我们表明，显着调整语言模型行为与小型手腕策划数据集是可行的。

有大量且不断增长的证据和文学探索人工智能（AI）技术对整个社会，政治和人类的影响。单独的平行工作已经探索了人类的存在风险，包括但不限于非对齐的人工通用智能（AGI）的风险。在本文中，我们认为当前和近期人工智能技术有可能通过充当中间风险因素来促进存在风险的观念，并且这种潜力不仅限于不规则的AGI场景。我们提出这样的假设，即AI的某些已经记录的影响可以充当存在的风险因素，从而放大了先前确定的存在风险来源的可能性。此外，即使在没有人工通用智能的情况下，未来十年的未来发展也有可能极大地加剧这些危险因素。我们的主要贡献是对潜在的AI风险因素以及它们之间的因果关系的（非排斥）的解释，重点是AI如何影响电力动态和信息安全。该博览会表明，从AI系统到没有假设未来AI能力的存在风险存在因果途径。

如何将新兴和全面的技术（例如AI）整合到我们社会的结构和运营中是当代政治，科学和公众辩论的问题。它从不同学科中产生了大量的国际学术文献。本文分析了有关人工智能调节（AI）的学术辩论。该系统审查包括在2016年1月1日至2020年12月31日之间发表的73份同行评审期刊文章样本。分析集中于社会风险和危害，监管责任问题以及可能基于风险的政策框架在内和基于原则的方法。主要利益是拟议的监管方法和工具。提出了各种形式的干预措施，例如禁令，批准，标准设定和披露。对所包括论文的评估​​表明该领域的复杂性，这表明其早产和剩余的缺乏清晰度。通过对学术辩论进行结构性分析，我们在经验和概念上均可更好地理解AI和监管的联系以及基本规范性决策。科学建议与拟议的欧洲AI调节的比较说明了调节的特定方法，其优势和缺点。

Nearly all jurisdictions in the United States require a professional license exam, commonly referred to as "the Bar Exam," as a precondition for law practice. To even sit for the exam, most jurisdictions require that an applicant completes at least seven years of post-secondary education, including three years at an accredited law school. In addition, most test-takers also undergo weeks to months of further, exam-specific preparation. Despite this significant investment of time and capital, approximately one in five test-takers still score under the rate required to pass the exam on their first try. In the face of a complex task that requires such depth of knowledge, what, then, should we expect of the state of the art in "AI?" In this research, we document our experimental evaluation of the performance of OpenAI's `text-davinci-003` model, often-referred to as GPT-3.5, on the multistate multiple choice (MBE) section of the exam. While we find no benefit in fine-tuning over GPT-3.5's zero-shot performance at the scale of our training data, we do find that hyperparameter optimization and prompt engineering positively impacted GPT-3.5's zero-shot performance. For best prompt and parameters, GPT-3.5 achieves a headline correct rate of 50.3% on a complete NCBE MBE practice exam, significantly in excess of the 25% baseline guessing rate, and performs at a passing rate for both Evidence and Torts. GPT-3.5's ranking of responses is also highly-correlated with correctness; its top two and top three choices are correct 71% and 88% of the time, respectively, indicating very strong non-entailment performance. While our ability to interpret these results is limited by nascent scientific understanding of LLMs and the proprietary nature of GPT, we believe that these results strongly suggest that an LLM will pass the MBE component of the Bar Exam in the near future.

Large language models have recently attracted significant attention due to their impressive performance on a variety of tasks. ChatGPT developed by OpenAI is one such implementation of a large, pre-trained language model that has gained immense popularity among early adopters, where certain users go to the extent of characterizing it as a disruptive technology in many domains. Understanding such early adopters' sentiments is important because it can provide insights into the potential success or failure of the technology, as well as its strengths and weaknesses. In this paper, we conduct a mixed-method study using 10,732 tweets from early ChatGPT users. We first use topic modelling to identify the main topics and then perform an in-depth qualitative sentiment analysis of each topic. Our results show that the majority of the early adopters have expressed overwhelmingly positive sentiments related to topics such as Disruptions to software development, Entertainment and exercising creativity. Only a limited percentage of users expressed concerns about issues such as the potential for misuse of ChatGPT, especially regarding topics such as Impact on educational aspects. We discuss these findings by providing specific examples for each topic and then detail implications related to addressing these concerns for both researchers and users.

Developing safe and useful general-purpose AI systems will require us to make progress on scalable oversight: the problem of supervising systems that potentially outperform us on most skills relevant to the task at hand. Empirical work on this problem is not straightforward, since we do not yet have systems that broadly exceed our abilities. This paper discusses one of the major ways we think about this problem, with a focus on how to turn it into one that can be productively studied empirically. We first present an experimental design centered on choosing tasks for which human specialists succeed but unaided humans and current general AI systems fail. We then present a proof-of-concept experiment following meant to demonstrate a key feature of this experimental design and show its viability with two question-answering tasks: MMLU and time-limited QuALITY. On these tasks, we find that human participants who interact with an unreliable large-language-model dialog assistant through chat -- a trivial baseline strategy for scalable oversight -- substantially outperform both the model alone and their own unaided performance. These results are an encouraging sign that scalable oversight will be tractable to study with present models and bolster recent findings that large language models can productively assist humans with difficult tasks.