An increasing number of jurisdictions now permit people to access voluntary assisted dying (VAD) to end their lives. Access to VAD is controlled by eligibility criteria, which vary in each jurisdiction meaning that not everyone who may wish to access VAD can. In this article, we focus on one specific cohort, namely people living with dementia, who may seek to access VAD on the basis of their dementia. We consider whether it may be possible for such individuals to access VAD in jurisdictions where VAD is lawful, and under what circumstances. To do this, we undertake an in-depth comparative analysis of all the relevant legal frameworks (as well as relevant policy and guidelines) to describe and categorize various VAD models where VAD for dementia may be possible. Those models include concurrent request models (where VAD is accessed and administered when a person has decision-making capacity) and advance request models (where VAD is requested when a person has decision-making capacity but accessed at a point when they no longer have such capacity). We conclude by considering the implications these models have for the availability of VAD for dementia based on eligibility, as well as potential practice implications.
Abstract Biometric identification and AI-driven emotion recognition are moving rapidly from the laboratory into the workplace, subjecting workers to continuous surveillance of their inner lives. This article examines whether the two central instruments of EU law governing algorithmic management—the General Data Protection Regulation (GDPR) and the Artificial Intelligence Act (AI Act)—can protect the worker as a human being rather than as a source of data. A doctrinal analysis of both instruments, together with the guidance of the Article 29 Working Party, the European Data Protection Board, and the European Commission’s Guidelines on Prohibited AI Practices, reveals structural gaps: the GDPR’s purpose-limitation and consent mechanisms malfunction in the employment context, while the AI Act’s prohibition of workplace emotion recognition rests on the technology’s unreliability rather than on a principled account of why workers’ inner lives deserve protection. The article argues that these fragmented protections should be connected by a unifying normative principle—the right to be human at work—grounded in Articles 1 and 31 of the EU Charter of Fundamental Rights, and proposes recognizing emotional data as a special category of data and introducing a right not to be emotionally profiled.
Abstract A new study by researchers at Stanford University describes the development of ‘xenocortical mice’: mice with a cortex comprised almost entirely of human-derived neurons. Their groundbreaking research has the potential to enable medical breakthroughs for a range of neurological and psychiatric disorders. The researchers offer detailed analysis of the neurophysiological and behavioral properties of these chimeric mice, comparing them with both wild-type control mice, and ‘apallial’ mice lacking most of their cortex. Here, we outline some key findings from the study and ask what these findings tell us about the capacity for consciousness in these mice, as well as their potential for human-like properties of cognition or conscious experience. Based on this discussion, we summarize the ethical and regulatory landscape relevant to chimeric animals and lay out an ethically-informed research roadmap to guide future research in this important area.
The US Food and Drug Administration (FDA) issues complete response letters (CRLs) after completing a review when it determines that it cannot approve a drug or biological product application in its current form. Historically, the FDA has treated CRLs as confidential, leaving sponsors largely in control of how non-approval decisions are described publicly. This approach has been criticized for relying on an expansive view of confidentiality, which can obscure safety, efficacy, and quality-related concerns, and for enabling selective or incomplete sponsor communications. In 2025, the FDA shifted course. On July 10, 2025, the FDA posted more than 200 CRLs tied to applications that had subsequently been approved, consolidating information that was often already embedded in the approval action packages. On September 4, 2025, the FDA released 89 previously unpublished CRLs for pending or withdrawn applications, and committed to real-time disclosure of newly issued CRLs going forward. Using a comparative FDA-European Medicines Agency (EMA) lens, we assess what these initiatives changed at launch and how early implementation through May 2026 affects their durability, legal footing, and information value. By May 2026, the FDA's CRL initiative had evolved into a functioning yet unsettled disclosure infrastructure: openFDA now offers searchable access to CRLs for approved and unapproved applications, while the FDA's FY 2027 budget request simultaneously seeks clearer statutory authority to disclose certain information in CRLs. We argue that FDA CRLs and EMA assessment outputs are complementary, and we propose lessons that FDA can learn from the EMA to maximize the value of CRL disclosure: codifying timelines and redaction standards, adding neutral decisional summaries where redactions or technical language would otherwise obscure the basis for non-approval, and standardizing the classification and discoverability of prior-cycle CRLs in approval packages. These reforms can protect legitimate confidential commercial information while enabling learning from unsuccessful applications.
Rapid scientific change is exposing the limits of the UK's regulation of human reproductive and developmental biology. Embryo-like structures, in vitro gametogenesis, organoids and heritable genome editing advance understanding of human development and physiology and support new applications in assisted reproduction and clinical care, yet the current regulatory framework is ill-equipped to govern them coherently. A case is made for coordinated, technology-neutral regulatory change. The article proposes the principles of a model UK Human Reproductive and Developmental Biotechnology Act and regulator, the Human Reproductive and Developmental Biotechnology Authority-or HRBA, pronounced 'harbor'-to provide a permissions route across research and translation. The model Act is a modular, risk-tiered scheme with consistent standards and tailored red lines. It is designed to accommodate boundary cases, to integrate and build upon the existing remit of the Human Fertilisation and Embryology Authority, and coordinate with other established regulators to retain institutional expertise and avoid unnecessary duplication. The proposal is ambitious and would help give effect to UK Government commitments to safely advance the life sciences in this area. But even if piecemeal amendments are ultimately favored, it clarifies the underlying choices and trade-offs and provides a benchmark against which changes can be tested.
