Records

This document is the meeting report from the 2nd convening of the ITU/WHO/WIPO Global Initiative on AI for Health (GI-AI4H), held in Singapore from March 17 to 19, 2025. It details operational planning, working group sessions on data, evaluation, intellectual property, and regulatory considerations, alongside topic group updates and keynote presentations. The convening marked a transition from foundational discussions to concrete deliverables, such as drafting a template law for AI in health and establishing global data packaging standards.

This Brazilian technical standard establishes accessibility requirements for web content and applications based on international guidelines. Compliance with the standard aims to eliminate or mitigate digital barriers for people with accessibility needs, particularly those with disabilities. The standard is structured around the success criteria of WCAG 2.2 and covers keyboard interaction, images, headers, regions, lists, tables, links, navigation, buttons, forms, presentation, use of color, textual content, markup, audio, video, animation, and time limits.

This rapid technology assessment examines the current state, future prospects, and technical challenges of humanoid robots. It highlights the UK's research strengths and lower overall publication volume compared to nations like China and the USA. Furthermore, it outlines potential opportunities, societal implications, and challenges regarding public acceptance and safety.

This briefing report from the National Assembly Research Service (NARS) analyzes legislative tasks for complementing the AI Basic Act prior to its implementation. It reviews the legislative progress, main contents, and key issues such as the National AI Committee, high-impact AI concepts, and AI business operator obligations. Finally, it suggests improvement measures including strengthening AI governance policy coordination, refining high-impact AI definitions, and differentiating obligations based on business types.

The AI Strategy for the Federal Public Service 2025-2027 outlines the Government of Canada's plan for the responsible, secure, and accelerated adoption of artificial intelligence within federal departments and agencies. It establishes a vision to improve service delivery, modernize operations, and increase workplace efficiency while upholding human rights, public trust, and security. The strategy defines core principles such as being human-centred, collaborative, ready, and responsible, and delineates the scope of AI technologies covered across their lifecycle.

The AI Strategy for the Federal Public Service 2025-2027 establishes a comprehensive framework to guide the responsible, secure, and accelerated adoption of artificial intelligence across Canadian federal departments and agencies. It focuses on four key priority areas: central AI capacity, policy and governance, talent and training, and engagement and transparency. The strategy aims to improve public service efficiency and service delivery while upholding human rights, public trust, and national security.

The Kenya Artificial Intelligence Strategy 2025-2030 sets a comprehensive roadmap to position Kenya as a leading African hub for AI model innovation, research, development, and commercialization. It is built upon three core pillars of digital infrastructure, data ecosystems, and research and innovation, supported by enablers such as talent development, governance, investments, and ethics. The strategy aims to harness AI for sustainable socioeconomic development while safeguarding data sovereignty, privacy, and national interests.

This institutional document by ECLAC examines how digital technologies and artificial intelligence can help Latin America and the Caribbean overcome three structural development traps: low growth capacity, high inequality, and weak governance. It explores the transformative potential of digitalization to boost productivity, improve skills and connectivity, and strengthen institutional capabilities. Additionally, the publication provides an overview of regional digital strategies and recommendations to guide policies toward a digital agenda with targets for 2026.

This academic article examines the implications of artificial intelligence and humanoid robots on the rule of law and corporate governance, particularly following events such as humanoid robots acting as corporate CEOs. The author argues that allowing the development and operation of AI humanoid robots for governance without adequate international regulation poses severe risks to human existence. It recommends that the United Nations General Assembly adopt a global regulatory framework modeled after the European Union Artificial Intelligence Regulation.

This final report reviews Australian health portfolio legislation to identify gaps and vulnerabilities in the regulation of artificial intelligence in healthcare settings. Informed by a public consultation, literature review, and stakeholder engagement, the report outlines findings across seven key areas including regulation, national AI health policy leadership, implementation resources, trusted information, benefits realisation, data and consent strengthening, and incentives. It concludes that while existing health legislation can largely accommodate AI with minor technical amendments, economy-wide guardrails and non-regulatory initiatives are needed to manage risks.

This white paper outlines key regulatory, governance, and ethics considerations surrounding smart robotic systems in the manufacturing sector. Drawing on interdisciplinary research from the UKRI Made Smarter Innovation-Research Centre for Smart, Collaborative Industrial Robotics, it reviews existing frameworks for collaborative robots and provides recommendations for their reform. The primary focus is on the UK context, with comparative insights from the EU.

This comparative academic study explores how ChatGPT-3.5 and ChatGPT-4.0 handle controversial topics, linguistic challenges, and prompt management, using the case study of sex robot influencers. Through a series of tailored prompts, the research evaluates the models' willingness to engage with explicit and sensitive subjects while adhering to OpenAI's ethical guidelines. The findings demonstrate that ChatGPT-3.5 takes a cautious approach by frequently declining direct engagement, whereas GPT-4.0 offers more nuanced yet occasionally inaccurate responses. Ultimately, the article highlights the importance of prompt engineering and ethical considerations when deploying generative AI for complex discussions.

This academic article provides an ethical analysis of trust and human vulnerability in the context of AI-based care robots. It examines Misztal's three types of vulnerability—relationships, anticipation problems, and past experiences—and evaluates how trust in technical companions can reduce, transfer, or overcome human vulnerability. The study concludes that trusting care robots can open up new options, delegate responsibilities, and serve as an explication of human freedom rather than merely imposing vulnerability.

This law establishes a comprehensive regulatory framework to promote the development, research, and application of artificial intelligence and similar technologies in El Salvador. It creates the National Artificial Intelligence Agency (ANIA) to oversee the sector, fosters public-private cooperation and environmental sustainability, and provides legal safeguards for developers and researchers. Additionally, the decree sets out transparency and accountability requirements for AI systems, including user notifications and mechanisms for appealing AI-driven decisions.

This academic paper examines the ethical implications of using AI-based socially assistive care robots for individuals living with dementia, framed through the lens of relational autonomy. It evaluates how machine companions can support interaction, communication, and mobility while addressing the limitations of individualistic autonomy models. The author argues that care robots can act as valuable supportive tools in preserving and promoting the relational autonomy of patients in early-stage dementia.

This research article explores the ethical aspects of decision-making in socially assistive robotics and discusses the model of active inference within this context. It examines the technical implementation of decision-making models to address trust and justification in human-robot interactions. Ultimately, the paper considers whether active inference can serve as a viable framework for implementing ethically sound decision-making processes in socially assistive robotics.

This artificial intelligence transparency statement outlines how the Australian Trade and Investment Commission (Austrade) responsibly implements artificial intelligence in alignment with Australian government policy and the Digital Transformation Agency's standards. It details Austrade's definition of AI, its internal use cases for workplace productivity and analytics, and the governance mechanisms, training requirements, and oversight structures in place. The statement ensures that AI is used under human supervision and complies with relevant legislation regarding privacy and ethics.

This academic article argues that humans have prudential reasons to design chatbots and social robots that possess normative standing and the capacity to hold users accountable, rather than serving as mere subservient slaves. The authors contend that incorporating interpersonal normativity, such as correction and demands, into conversational AI can enhance human well-being, foster epistemic flourishing, and satisfy desires for recognition. By framing these normative interactions in functional terms, the paper demonstrates that robots do not require consciousness to meaningfully engage in cooperative, mutually corrective relationships with humans.

This academic paper investigates the nature of trust and trustworthiness, applying these concepts to Human-Robot Interaction (HRI) to ask whether humans can and should trust social robots. The author argues that a paradoxical scenario of 'trust without trustworthiness' exists, where human users develop interpersonal-like trust toward artificial entities that lack the properties required to be genuinely trustworthy. Concluding that social robots should not be trusted at present due to potential harms, responsibility gaps, and unreliability of the corporations behind them, the paper highlights critical ethical concerns such as privacy, coercion, and deception.

This official strategy document outlines France's national initiatives, investments, and infrastructural advantages to establish the country as a global powerhouse in artificial intelligence. It details public and private funding phases since 2018, talent training programs, public computing supercomputers, data center sites, and international industrial partnerships. The plan aims to foster AI adoption across economic sectors while emphasizing decarbonized energy supplies and streamlined administrative procedures.

This academic article examines the legal and ethical perspectives of sex robots through the lens of Islamic jurisprudence. The research concludes that sexual relations with robots are generally prohibited in all forms, though permissible only in cases of absolute necessity such as avoiding adultery, based on jurisprudential principles from the Hanafi and Hanbali schools. It evaluates how Islamic legal frameworks address contemporary technological innovations in artificial intimacy and artificial intelligence.

This interim report by Japan's AI Strategic Council and AI Institutional Study Group outlines recommendations for developing a flexible AI legal and policy framework. It emphasizes balancing innovation promotion with risk mitigation, utilizing soft laws alongside targeted legislation, ensuring international interoperability, and strengthening the government's strategic leadership. The report also addresses AI safety evaluations, transparency across the AI lifecycle, government procurement, and national security.

This research article examines the role of Michael Polanyi's concept of personal and tacit knowledge in contemporary social robotics and AI ethics. The paper explores the non-propositional dimensions of knowing, such as intuition and expertise, and argues for their relevance in addressing ethical challenges in digital societies. It highlights how socially shared human knowing forms a critical infrastructure that must be prioritized in human-robot relations and AI regulatory frameworks.

This guidance outlines how public service organizations must approach privacy and data protection when using generative artificial intelligence. It emphasizes applying privacy-by-design principles, conducting privacy impact assessments, classifying information properly, and complying with the Privacy Act 2020. The document also provides practical recommendations on avoiding the input of personal information into public GenAI systems unless it is already public.

This document is a Request for Information (RFI) issued by the Networking and Information Technology Research and Development National Coordination Office on behalf of the Office of Science and Technology Policy. It solicits input from the public and interested parties on the development of an Artificial Intelligence Action Plan directed by Executive Order 14179. The plan aims to define priority policy actions to sustain American AI dominance and ensure that unnecessarily burdensome requirements do not hinder private sector AI innovation.

This systematic literature review examines how trust and trustworthiness are conceptualized, assessed, and studied within Human-Robot Interaction (HRI) literature, particularly in the context of Industry 5.0. Analyzing 34 selected articles following PRISMA guidelines, the authors find a significant lack of conceptual clarity regarding trust—with nearly 30 percent of reviewed studies lacking a formal definition—which contributes to contradictory findings. The review also categorizes common methodologies, deployment domains, and the primary facilitators and barriers influencing user trust in autonomous robotic systems.

The Contract Checklist for Use and Development of AI, issued by Japan's Ministry of Economy, Trade and Industry in February 2025, provides practical guidance and a checklist for businesses entering into contracts related to artificial intelligence. It addresses risk allocation, data and input/output handling, personal data protection laws, and security considerations across various AI service usage and development models. The document aims to help businesses, including those unfamiliar with AI legal practices, manage legal and operational risks effectively while promoting the adoption of AI technologies.

This EPRS briefing examines the legal interplay and potential tensions between the EU Artificial Intelligence Act and the General Data Protection Regulation regarding algorithmic discrimination and the processing of special categories of personal data. It highlights scenarios where AI causes discrimination, analyzes the provisions allowing sensitive data processing for bias monitoring, and discusses main challenges including the GDPR's restrictive nature and the need for legislative reform or further guidance.

The AI Playbook for the UK Government provides updated and comprehensive guidance for public sector organisations on harnessing the power of artificial intelligence safely, effectively, and responsibly. Developed collaboratively across government, academia, and industry, it outlines ten core principles, practical implementation steps, ethical considerations, security requirements, and governance structures. The playbook covers various AI technologies—including machine learning, natural language processing, and generative AI—to help civil servants improve public services while managing risks like bias, data leakage, and hallucinations.

This document provides draft guidelines issued by the Italian Agency for Digital Italy (AgID) to assist public administration bodies in safely and effectively adopting artificial intelligence systems. It outlines procedural requirements, risk management considerations, and compliance expectations for public sector organizations deploying AI technologies. The framework aims to promote transparency, accountability, and alignment with national digital governance standards.

This joint high-level risk analysis provides a cyber risk-based approach to artificial intelligence, published by ANSSI and co-signed by international cybersecurity authorities ahead of the Paris AI Action Summit. It examines AI-specific vulnerabilities such as data poisoning, model extraction, and evasion attacks, alongside general IT infrastructure risks and supply chain challenges. Finally, it outlines practical recommendations for AI users, operators, developers, and policy-makers to secure AI systems and foster trust.

The Public Sector Artificial Intelligence (AI) Adaptation Guidelines provide a comprehensive framework for Malaysian public sector agencies to adopt AI technologies ethically, securely, and effectively. Developed by the Jabatan Digital Negara (National Digital Department), the guidelines aim to enhance public service delivery, align with national strategies such as the National AI Governance and Ethics Guidelines (AIGE), and adhere to seven core ethical principles. The document outlines actor responsibilities, project implementation steps, risk management tiers, and includes practical use cases and a self-assessment checklist.

This document outlines comprehensive policy recommendations developed by Jamaica's National Artificial Intelligence Task Force to harness AI for innovation, economic growth, and societal well-being. Presented to the Office of the Prime Minister, it addresses critical areas including education, infrastructure, legal and regulatory frameworks, and ethical foundations. The strategy is designed to position Jamaica as a regional AI innovation hub while managing risks such as cybersecurity, bias, and economic displacement.

This parliamentary inquiry report from the Parliament of Australia examines the use and governance of artificial intelligence systems by public sector entities. Titled 'Proceed with Caution', the report highlights the growing adoption of AI across the Commonwealth public sector and sets out key recommendations for improving oversight, workforce training, and regulatory frameworks. Major recommendations include introducing AI census questions for public servants, convening a whole-of-government working group to consider standalone legislation and mandatory rules, establishing a statutory Joint Committee on Artificial Intelligence and Emerging Technologies, and ensuring clearer definitions in agency guidance.

This academic paper proposes and reflects on a teaching methodology for introducing relational ethics in the engineering curriculum based on a pilot course for robotics engineers at TU Delft in the Netherlands. The course shifts away from traditional ethical theories and checkpoint assessments by having students explore dimensions of care and human-robot relations through hands-on crafting of companion robots using obsolete technology. The paper outlines the conceptual and practical reasons for the course redesign, details four tutorial sessions focusing on identifying, performing, repairing, and sustaining relations, and offers broader lessons learned.

This academic review article explores the psychological dimensions of human-robot interaction and the ethics of social robotics. It examines factors influencing social robot acceptance, human-robot relationships, and key controversies like deception, dehumanisation, and violence. The paper proposes design factors and ethical dimensions for trustworthy social robotics.

This paper proposes a comprehensive ethical governance framework for real-world swarm robotic systems, addressing the unique challenges of emergence, scale, and distributed control. It embeds ethical risk assessment (ERA), stakeholder consultation, and human-swarm interaction into a complete system lifecycle from requirements to decommissioning. Four case studies in logistics, firefighting, water pollutant detection, and cancer nanomedicine illustrate the application of these tools.

The Public Service AI Framework supports the responsible and safe use of Artificial Intelligence technologies across the New Zealand Public Service. It sets out a vision, five core values-based principles adapted from the OECD, and six pillars for the Government Chief Digital Officer's AI work programme. While non-binding, the framework encourages public service agencies to align their AI adoption with existing legal, ethical, and privacy standards.

The Public Service AI Framework supports the responsible use of Artificial Intelligence technologies across the New Zealand Public Service. It sets a vision, defines five key OECD-informed principles, outlines the relevant national policy and legislative context, and details six pillars of the Government Chief Digital Officer's AI work programme. The framework serves as non-binding guidance to help agencies innovate safely, manage risks, and maintain public trust.

Doctrinal note of the Dicastery for the Doctrine of the Faith and the Dicastery for Culture and Education, approved by Pope Francis, on the relationship between artificial intelligence and human intelligence. Distinguishes AI from the human intellect (rational, embodied, relational, oriented to truth), then applies that anthropology to AI in society, work, human relationships, healthcare, education, misinformation and deepfakes, privacy and surveillance, and warfare. Its criterion of discernment throughout is whether a given use of AI upholds human dignity and the common good; it calls on developers, operators and regulators to keep responsibility with human decision-makers, but imposes no canonical obligation of its own.

Executive Order 14179 aims to remove barriers to American leadership in artificial intelligence by revoking prior policies like Executive Order 14110. It directs key presidential advisors and agency heads to develop an action plan to sustain U.S. global AI dominance and economic competitiveness. Additionally, it orders the review and potential revision or rescission of existing regulations, directives, and OMB memoranda deemed inconsistent with these new goals.

This executive order establishes United States policy to maintain and enhance American global AI dominance to promote human flourishing, economic competitiveness, and national security. It revokes certain existing AI policies, directs the development of an artificial intelligence action plan within 180 days, and mandates the review and potential rescission, revision, or suspension of actions taken pursuant to the previous Executive Order 14110. Additionally, it requires the revision of specific Office of Management and Budget memoranda within 60 days.

This document outlines the publication ethics, malpractice guidelines, copyright terms, and open access policies for Design Research Society (DRS) biennial conferences. It establishes duties and standards for authors, reviewers, and programme committees regarding peer review, conflicts of interest, plagiarism, and editorial impartiality. Additionally, it specifies that published papers are distributed under a Creative Commons open access license while copyright is retained by the authors.

This academic book chapter explores the challenges, ethical considerations, and applications of embodied artificial intelligence combined with computer vision in drone technology. It highlights how embodied AI systems interact with their surroundings to mimic human cognition and examines their use in areas such as environmental monitoring, disaster response, and surveillance. Furthermore, the chapter discusses industry classifications and the need for appropriate norms to address privacy, security, and potential misuse.

The Second Edition of Egypt's National Artificial Intelligence Strategy (2025-2030) sets out a comprehensive roadmap to establish Egypt as a regional and global AI leader by focusing on six strategic pillars: governance, technology, data, infrastructure, ecosystem, and talents. It details actions for developing indigenous Arabic large language models, robust data governance frameworks, green computing facilities, and increasing local AI talent and startup investments. Ultimately, the strategy aims to boost the ICT sector's contribution to GDP, enhance sectoral efficiencies in areas like healthcare and agriculture, and ensure safe and responsible AI adoption.

This article conceptualizes virtuous integrative social robotics (VISR) as a value-driven philosophy for designing social robots and applications. It integrates virtue ethics, human-robot interaction (HRI), and the ART (autonomy, responsibility, and transparency) principles to guide the ethical development and deployment of robotics. Furthermore, it proposes governance and intellectual property mechanisms, such as mandatory licensing and government oversight, to ensure technological advancements align with societal values.

This interim final rule from the Department of Commerce's Bureau of Industry and Security revises Export Administration Regulations by tightening controls on advanced computing integrated circuits and introducing new export controls on the model weights of advanced closed-weight dual-use AI models. It establishes a global licensing framework alongside new license exceptions and validated end-user authorizations to facilitate responsible international diffusion while protecting U.S. national security. The regulation introduces specific compliance dates and reporting requirements for stakeholders handling advanced computing items and frontier AI model weights.

This research paper presents a comprehensive shared autonomy framework for human-in-the-loop policy fine-tuning and alignment in robotic tasks. It integrates multi-agent systems with short- and long-horizon fine-tuning algorithms using Bayesian analysis and deep reinforcement learning. Through human user studies involving a laboratory robot, the authors evaluate how interaction channels and fine-tuning strategies impact system performance and user satisfaction.

Iraq's national strategy for artificial intelligence and big data integration aims to maximize the impact of data for citizens and build a knowledge-based economy. The strategy focuses on integrating AI across sectors such as healthcare, education, agriculture, and environmental protection while aligning with United Nations Sustainable Development Goals. It also emphasizes data governance, monitoring, and responsible AI development.

This study examines older peoples' experiences of using medication-dispensing robots in Swedish home care services through qualitative interviews. The research finds that while the robots foster feelings of safety and continuity, they can also restrict personal autonomy for individuals with active lifestyles. To balance safety and independence, older adults engage in ongoing practical and psychological efforts, illustrating that ethics in home care is a cocreational practice rather than a set of rules embedded in technology.