Regulation Of AI Research Articles

Evolving cybersecurity of AI-featured digital products and services: Rise of standardisation and certification?

The field of cybersecurity has changed dramatically since the Cybersecurity Strategy for the Digital Decade was presented by the European Commission and the High Representative of the Union for Foreign Affairs and Security Policy in December 2020. The Cybersecurity Strategy highlights the potential of AI as a new technology, but also the need for cyber security of AI technology. Indeed, since the strategy was adopted, AI has shown that it has enormous potential for growth, but also several risks and vulnerabilities that this new technology brings. The paper analyses the shift and further development in the field of cybersecurity of digital products and services, AI itself as a technology, as well as products and services that will contain an AI component. In our opinion, the way to ensure that not only AI technology itself, but also products and services are cyber-secure, is to achieve a high level of standardisation of best practices, as there are many gaps in this area. The adoption of technical standards will fully form a path for conformity assessment and certification of not only AI systems but also AI-featured digital products and services. However, the current regulatory trend is to adopt a comprehensive legal regulation of AI even before such technical standards are fully developed and adopted. We consider this risky. Despite the well-intentioned effort to define and regulate AI, the purpose set forth in the AIA may not be achieved, as the requirements adopted in this way can very quickly become unnecessarily burdensome or even outdated due to increasing technological development. The proof of this is also the recent rise of large ML models, known as foundation models, which significantly changed the previous understanding of the creation of AI systems. It will be the technological development of AI, AI specific standardisation, and subsequent certification of digital products and services, which will govern future activities in building Europe's cyber resilience.

Directors’ duties and AI regulation

ABSTRACT Companies are powerful initiators and adopters of AI, as they have been of previous industrial revolutions. Their boards of directors and the regulatory frameworks that govern directors will inevitably influence the development and implementation of AI. However, despite intense attention to AI regulation the regulatory contribution of directors’ duties remains relatively unexamined. This article critically analyses the capacity of these longstanding legal principles to assist with wider AI regulatory goals, demonstrating that directors’ duties have direct consequences for the regulation of AI. Evolved in response to the governance context of a complex artificial entity, the corporation, directors’ duties map onto key AI regulatory requirements. These include the need for broadly accepted but flexible and responsive norms, and a capacity to engage post-facto with both nuanced, complex fact situations and opaque decision making processes. Directors’ duties jurisprudence brings with it experience of reverse-engineering fact situations and responding to disembodied legal actors, two factors that may have wider AI regulatory application. Directors’ duties also demonstrate responsiveness to changing social perspectives, facilitating integration of developing attitudes to AI. Attending to the AI regulatory role played by directors’ duties can inform effective AI regulation and facilitate improved regulatory design overall.

Towards a Human Rights-Based Approach to Ethical AI Governance in Europe

As AI-driven solutions continue to revolutionise the tech industry, scholars have rightly cautioned about the risks of ‘ethics washing’. In this paper, we make a case for adopting a human rights-based ethical framework for regulating AI. We argue that human rights frameworks can be regarded as the common denominator between law and ethics and have a crucial role to play in the ethics-based legal governance of AI. This article examines the extent to which human rights-based regulation has been achieved in the primary example of legislation regulating AI governance, i.e., the EU AI Act 2024/1689. While the AI Act has a firm commitment to protect human rights, which in the EU legal order have been given expression in the Charter of Fundamental Rights, we argue that this alone does not contain adequate guarantees for enforcing some of these rights. This is because issues such as EU competence and the principle of subsidiarity make the idea of protection of fundamental rights by the EU rather than national constitutions controversial. However, we argue that human rights-based, ethical regulation of AI in the EU could be achieved through contextualisation within a values-based framing. In this context, we explore what are termed ‘European values’, which are values on which the EU was founded, notably Article 2 TEU, and consider the extent to which these could provide an interpretative framework to support effective regulation of AI and avoid ‘ethics washing’.

Government regulation or industry self-regulation of AI? Investigating the relationships between uncertainty avoidance, people’s AI risk perceptions, and their regulatory preferences in Europe

Government regulation or industry self-regulation of AI? Investigating the relationships between uncertainty avoidance, people’s AI risk perceptions, and their regulatory preferences in Europe

Artificial intelligence integration in healthcare: perspectives and trends in a survey of U.S. health system leaders

BackgroundThe healthcare sector is rapidly integrating artificial intelligence-derived predictive models (AIDPM) to enhance clinical decision support, operational efficiency, and patient experiences. However, research on management strategies for AIDPM acquisition, deployment, and governance remains limited. This study examines changes in AIDPM integration and governance since 2021, with a particular focus on large language models and health equity considerations.ResultsOur survey of health system leaders achieved a 49% response rate (32/65). While 84% of institutions reported using AIDPM in clinical practice, only 53% had established dedicated teams for these models. Compared to 2021, there was a significant increase in representation from experts in clinical informatics, operations, and quality improvement on AIDPM teams. Most organizations (41%) primarily purchased AIDPM from external vendors. Support for integrating large language models into healthcare practices was unanimous among respondents. The principal obstacles to AIDPM adoption included regulatory concerns, data security, workflow integration, and clinician acceptance. A large majority (72%) of respondents supported government regulation of AI in healthcare. While 76% of organizations reported having a team member dedicated to health equity, ethicists and diversity leaders were underrepresented on AIDPM teams (18%). Organizations reported various efforts to promote health equity, but involvement of frontline clinicians in AIDPM development and its impact on health equity was significantly less common.ConclusionsClinical adoption of AIDPM faces challenges due to the absence of established best practices. Health system leaders strongly support federal regulations for AI in healthcare. These regulations could provide quality and safety standards. The study highlights the need for developing evaluation guidelines, especially for large language models. It also reveals a lack of uniform involvement of frontline clinicians and equity experts in AIDPM governance. Their involvement could increase adoption and trust of these new AI tools. Future research should assess healthcare systems' adherence to emerging regulations and best practice frameworks. This research should emphasize patient safety and health equity. These findings underscore the urgent need for a comprehensive roadmap. This roadmap would guide the responsible implementation of AIDPM in healthcare settings.

Drone Regulation and AI Law: Assessing the Intersection of the EU Legal Frameworks for Unmanned Aircraft and Artificial Intelligence

Drones are becoming a more familiar sight in the skies. This growth has potential implications for all aspects of society. To ensure safe, secure and sustainable drone activities, regulatory authorities have been revoking, amending or introducing laws. The primary focus has been on aviation safety rules for the design, manufacture and operation of these unmanned aircraft. These rules do not directly address other domains. This paper assesses how the EU’s drone-specific safety-based rules interact with the EU’s AI legal framework that is currently being developed. While drone rules do not address AI directly, they do indirectly as, for example, autonomous drone operations are permitted in the single European sky. The general AI legal framework also applies to drone activities, despite such not being mentioned. This paper brings together the sector-specific EU drone rules and the EU’s generalAI legal framework to understand what the term ‘AI’ means within the context of drone operations, and assesses whether there are any inconsistencies, contradictions, overlaps or lacunae. This will determine the applicability and suitability of applying the general AI legal framework to a specific technology (drones) and sector (aviation), and whether the drone rules can contribute to the regulation of AI more generally.

What’s Your Stake in Sustainability of AI?: An Informed Insider’s Guide

It's no secret that AI systems come with a significant environmental cost. This raises the question: What are the roles and responsibilities of computing professionals regarding the sustainability of AI? Informed by a year-long informal literature review on the subject, we employ stakeholder identification, analysis, and mapping to highlight the complex and interconnected roles that five major stakeholder groups (industry, practitioners, regulatory, advocacy, and the general public) play in the sustainability of AI. Swapping the traditional final step of stakeholder methods (stakeholder engagement) for entanglement, we demonstrate the inherent entwinement of choices made with regard to the development and maintenance of AI systems and the people who impact (or are impacted by) these choices. This entanglement should be understood as a system of human and non-human agents, with the implications of each choice ricocheting into the use of natural resources and climate implications. We argue that computing professionals (AI-focused or not) may belong to multiple stakeholder groups, and that we all have multiple roles to play in the sustainability of AI. Further, we argue that the nature of regulation in this domain will look unlike others in environmental preservation (e.g., legislation around water contaminants). As a result, we call for ongoing, flexible bodies and policies to move towards the regulation of AI from a sustainability angle, as well as suggest ways in which individual computing professionals can contribute to fighting the environmental and climate effects of AI.

Ethics and Regulation of AI in Precision Oncology

Ethics and Regulation of AI in Precision Oncology

Large Language Model Use in Radiology Residency Applications: Unwelcomed but Inevitable

ObjectiveThis study explores radiology program directors’ perspectives on the impact of large language model (LLM) use among residency applicants to craft personal statements. MethodsEight program directors from the Radiology Residency Education Research Alliance participated in a mixed-methods study, which included a survey regarding impressions of artificial intelligence (AI)-generated personal statements and focus group discussions (July 2023). Each director reviewed four personal statement variations for five applicants, anonymized to author type: the original and three Chat Generative Pre-trained Transformer-4.0 (GPT) versions generated with varying prompts, aggregated for analysis. A 5-point Likert scale surveyed the writing quality, including voice, clarity, engagement, organization, and perceived origin of each statement. An experienced qualitative researcher facilitated focus group discussions. Data analysis was performed using a rapid analytic approach with a coding template capturing key areas related to residency applications. ResultsGPT-generated statement ratings were more often average or worse in quality (56%, 268 of 475) than ratings of human-authored statements (29%, 45 of 160). Although reviewers were not confident in their ability to distinguish the origin of personal statements, they did so reliably and consistently, identifying the human-authored personal statements at 95% (38 of 40) as probably or definitely original. Focus group discussions highlighted the inevitable use of AI in crafting personal statements and concerns about its impact on the authenticity and the value of the personal statement in residency selections. Program directors were divided on the appropriate use and regulation of AI. DiscussionRadiology residency program directors rated LLM-generated personal statements as lower in quality and expressed concern about the loss of the applicant’s voice but acknowledged the inevitability of increased AI use in the generation of application statements.

Implications of the novel EU AI Act for neurotechnologies

The EU AI Act, the first comprehensive regulation of AI, came into effect in August. Here, we provide an overview of the provisions that apply to the field of neurotechnology with respect to research and development and neuroscience practice and discuss some implications for the future.

A High-Level Structured Methodology for Development of AI Systems in Africa

AI is a potential game changer for Africa to address the specific challenges she faces in sectors like healthcare, climate change and water-related issues. However, the regulation of AI is still largely underdeveloped in Africa with some existing policies and frameworks still being young. Therefore, as the adoption of AI systems spreads across Africa, so does the need for a structured methodology to guide organizations in either developing new AI systems or onboarding existing ones while maintaining the quality and ethicality of these systems. This paper aims to develop a holistic methodology that provides comprehensive guidance to companies considering to develop new AI systems or onboard existing systems. The goal is to support the development and deployment of AI systems tailored to the specific needs of Africa. The proposed methodology employs a lifecycle approach that integrates both Agile and Waterfall frameworks. By combining the adaptive flexibility of Agile with the structured progression of Waterfall, this methodology ensures adaptability and thoroughness throughout the AI system's development and implementation phases. The integration of these methodologies offers a robust, adaptable framework that can be tailored to the unique demands of AI projects in Africa, from design to implementation, deployment as well as maintenance phases, thereby maximizing the potential impact of AI technologies in the region.

The Regulation of Generative AI: Lessons Learned from ChatGPT’s Suspension in Italy

The Regulation of Generative AI: Lessons Learned from ChatGPT’s Suspension in Italy

P-340 Reconciling diversity in health and genomic data collection with the regulation of AI in clinical genomics

Abstract Study question How far the existing and emerging EU regulations address concerns about lack of diversity in genomic databases which used to train AI in genomics? Summary answer The main regulations in the European Union provide no clear answers for reconciling the need for more diverse genetic data with regulatory standards. What is known already AI researchers need high quality genomic data that reflect the underlying diversity of human populations to develop and validate safe, effective, and equitable tools for genomic medicine. The problem is that most of the genomic data that are available to develop such tools are of insufficient quality, due to inappropriate use of population descriptors, and over representative of individuals of European ancestry. We used the example of developing algorithms to calculate polygenic risk scores ( PRS) and concerns about their accuracy for various sub-populations to illustrate our discussion. Study design, size, duration We have analysed the relevant EU regulations including, the upcoming AI Act, Medical Devices Regulation ( MDR), In Vitro Diagnostic Medical Devices ( IVDR) and the GDPR to assess how the concerns related to lack of diversity in genetic/genomic databases have been addressed. Participants/materials, setting, methods NA. Main results and the role of chance The main regulations for clinical AI in the European Union, namely those regulating medical devices, provide no clear answers for reconciling the need for more diverse genetic data with regulatory standards. On the other hand, some of the data protection regulation requirements arising from the EU General Data Protection Regulation (GDPR) may further challenge collection of sensitive attributes such as race/ancestry which might be needed for brining diversity to databases. On the other hand, several interesting impending new regulations addressing AI and algorithms such as the EU’s AI Act may help in some ways, but still fail to ultimately grapple with the fundamental obstacles in producing the diverse genomic data needed to achieve regulatory objectives. In particular, the proposed requirements related to data quality and transparency need to be accompanied by further guidance on how to define representativeness and how to address long-term impact on fairness concerns in the context of medical AI, when the developed tools are not of optimal use for populations not descending from European ancestry. Limitations, reasons for caution NA. Wider implications of the findings The legal analysis provided in this paper will provide insights for broader types of AI tools which may be implemented in the context of Assisted Reproductive Technologies and beyond. Trial registration number NA

Regulation of Artificial Intelligence and Penal Factors: Legal and Ethical Challenges

This bibliometric review analyzes the scientific production related to the variables Artificial Intelligence, Regulation and Ethics registered in the Scopus database during the period 2018-2023. The main objective of the study was to identify and characterize the volume of publications, achieving a total of 552 documents. The information collected was organized by graphs, categorizing it by Year of Publication, Country of Origin, Area of Knowledge and Type of Publication. The results reveal that the United States is the country with the highest number of publications, reaching a total of 99 scientific papers. The area of Computer Science stood out as the most prolific in terms of bibliographic contribution, with 236 documents. Likewise, Journal Articles represented 50% of total publications. This analysis also includes a qualitative study on the positions of various authors in relation to the topics addressed, providing a comprehensive view of the current state of research in this field. Among the main conclusions, it is stated that the regulation of AI is an evolving field that requires collaboration between legislators, technologists, ethicists and society in general, and managing flexible regulations is completely necessary taking into account the changing and constant evolution of everything related to AI today.

Сomparative legal study of artificial intelligence positioning in judicial proceedings.

The author studies the experience of using artificial intelligence in the framework of legal proceedings in the USA, China and Russia. This analysis includes not only an analysis of judicial practice, but also a study of strategic documents assigned at the government level of these countries. Goal of the research is to identify effective starting points for the implementation of AI in the judiciary, to determine the features of regulation of AI in the judicial process and specify prospects for determining the full autonomy of AI for judicial proceedings. The objectives of the study are to determine the main trends in the implementation of AI in the field of judicial process, highlight areas that can currently be algorithmically processed by AI, as well as determine future designs of the legal personality of AI, taking into account the opinion of the professional judicial community of Russia. The research methodology includes systemic, structural-functional, hermeneutic and comparative legal methods. The authors makes general conclusions about the prospects of AI for the judicial process and formulates optimal tracks for the use of AI in judicial cases, taking into account the comparative legal research conducted. Based on the comparative legal research carried out, the author formulated the increasing role of artificial intelligence technology used in the judicial process. The need to improve the legislative framework taking into account the technological specifics of AI and ensure an adequate level of information security when using this technology is noted. In addition, a starting point has been identified for the effective and safe implementation of AI as an autonomous participant in the process in relation to disputes that are considered in a simplified manner or contain a high proportion of data that is reliably verified and stored electronically.

A Model for Calculating the Indirect Added Value of AI for Business

The new world, where artificial intelligence is applied in every industry and aspect of life, is already becoming second nature to us. Strengthen the penetration of all the company’s business operations to withstand competitive pressures. Its impact is also apparent. The increasing importance of the added value of your application is being widely discussed, yet there is little discussion on how to measure it. This essay seeks to address this disparity. It raises the question of how to determine the indirect added value of artificial intelligence in the businesses that utilize it. In order to achieve this, we propose a model based on three main factor groups: cognitive elements, which influence how AI is perceived; behavioral elements, which affect behavior when using AI; and environmental variables, which impact laws governing the adoption or regulation of AI. The indirect value added by the application of AI is calculated by the organization’s management using numerical expressions for each set of elements.

Do Cases Generate Bad AI Law?

There’s an AI governance problem, but it’s not (just) the one you think. The problem is that our judicial system is already regulating the deployment of AI systems—yet we are not coding what is happening in the courts as privately driven AI regulation. That’s a mistake. AI lawsuits here and now are determining who gets to seek redress for AI injuries; when and where emerging claims are resolved; what is understood as a cognizable AI harm (and what is not), and why that is so. This Essay exposes how our judicial system is regulating AI today and critically assesses the governance stakes. When we do not adequately recognize how the generative AI cases being decided by today’s judges are already operating as a type of AI regulation, we fail to consider which emerging tendencies of adjudication about AI are likely to make good or bad AI law. For instance, litigation may do good agenda-setting and deliberative work as well as surface important information about the operation of private AI systems. But adjudication of AI issues can be bad too, given the risk of overgeneralization from particularized facts; the potential for too much homogeneity in the location of lawsuits and the kinds of litigants; and the existence of fundamental tensions between social concerns and current legal precedents. If we overlook these dynamics, we risk missing a vital lesson: AI governance requires better accounting for the interactive relationship between regulation of AI through the judicial system and more traditional public regulation of AI. Shifting our perspective creates space to consider new AI governance possibilities. For instance, litigation incentives (such as motivations for bringing a lawsuit or motivations to settle) or the types of remedies available may open up or close down further regulatory development. This shift in perspective also allows us to see how considerations that on their face have nothing to do with AI—such as access to justice measures and the role of judicial minimalism—in fact shape the path of AI regulation through the courts. Today’s AI lawsuits provide an early opportunity to expand AI governance toolkits and to understand AI adjudication and public regulation as complementary regulatory approaches. We should not throw away our shot.

Overview, Concepts, Viewpoints to the Possibility of Legal Regulation of Artificial Intelligence Technology

The purpose of the study is to disclose the concept of "artificial intelligence", to outline the possibilities of legal regulation of AI. The article considers the studies of various authors, among which it is necessary to name P.M. Morhat, A. Romaglio, V.L. Entin. The scientific novelty of the work lies in the fact that the category of "artificial intelligence" is considered from the position of the legislation of the Russian Federation, as well as based on the opinion of domestic and foreign scientists. As a result, based on the comparative analysis of information contained in the works of various authors, regulatory legal acts of the Russian Federation, foreign judicial practice, it is concluded that representatives of the legal and scientific communities demonstrate the lack of consensus on the question of whether the units of artificial intelligence and neural networks can be subjects of copyright. A. Ramalho is convinced that AI units cannot be subjects of copyright, which is confirmed by the analysis of the following court decisions: Decision of the Court of Appeal of the Federal Court of Australia dated 13.04.2022, Decision of Shenzhen Nanshan District People's Court dated 24 December 2019. The opposing view is reflected in the Japanese "Intellectual Property Strategy Program 2016", which uses the following argument: it seems impossible to distinguish the product of AI units from human intellectual products. It is impossible to agree with this, as the images generated by neural networks often cause the "uncanny valley" effect in people; no neural network has been able to pass the Turing test so far.

Regulatory Analogies, LLMs, and Generative AI

Abstract With the release of large language models such as GPT-4, the push for regulation of artificial intelligence has accelerated the world over. Proponents of different regulatory strategies argue that AI systems should be regulated like nuclear weapons posing catastrophic risk (especially at the frontiers of technical capability); like consumer products posing a range of risks for the user; like pharmaceuticals requiring a robust prerelease regulatory apparatus; and/or like environmental pollution to which the law responds with a variety of tools. This thinkpiece outlines the shape and limitations of particular analogies proposed for use in the regulation of AI, suggesting that AI law and policy will undoubtedly have to borrow from many precedents without committing to any single one.

Artificial Intelligence and Scientific Research:Prospects and Risks———Synthesis of the session “Artificial Intelligence and Paradigm Change in Science and Technology Innovation” in Tianjin Forum 2023

In the sub-forum “Transformation of Artificial Intelligence and Paradigm in Science and Technology Innovation” of Tianjin Forum 2023, scholars from both China and abroad discussed the impact of artificial intelligence. With the deepening development of the new technological revolution and industrial transformation, new-generation artificial intelligence technology is continuously making breakthroughs in research and application. AI not only promotes the transformation of material productivity, but also gradually emerges as a critical engine for enhancing the knowledge productivity. Since the emergence of the concept of "AI for Science", it has become an obvious proposition generally accepted by the academic community for its tremendous enabling capabilities for knowledge production. Artificial intelligence technology continues to achieve breakthroughs and gain widespread infiltration into the scientific research field, introducing new elements and momentum into scientific research and significantly catalyzing the enhancement of scientific research efficiency and paradigm shifts. AI-driven scientific research has become a new frontier in the global application of artificial intelligence. However, as artificial intelligence triggers paradigm shifts in social science research, the issues of data security, ethics, and value alignment that it brings about need to draw attention from the social science community. Scholars participating in the forum generally concurred that, in regulating the enabling role of AI in scientific research, it is imperative to take into account the specificities of various disciplines and stages, and comprehensively reasonable rational risk allocation mechanisms, platform. support mechanisms, and collaborative participation frameworks to achieve prudent, agile, and full lifecycle regulation of AI for Science.