Bioethics

Does brain-computer interface-based mind reading threaten mental privacy? ethical reflections from interviews with Chinese experts
Fangxu Han, Haidan Chen

Abstract:

Background: The rapid development of brain-computer interface (BCI) technology has sparked profound debates about the right to privacy, particularly concerning its potential to enable mind reading. While scholars have proposed the establishment of neurorights to safeguard mental privacy, questions remain about whether BCIs can genuinely decode inner thoughts and what makes their ethical implications distinctive.

Methods: This study conducted semi-structured interviews with 20 Chinese experts in the BCI and neuroscience fields to explore their perspectives on the concept, feasibility, and limitations of BCI-based mind reading (BMR). The transcriptions of the interviews were analyzed through reflexive thematic analysis to identify key themes and insights.

Results: The findings reveal a range of expert perspectives on the interpretations and feasibility of BMR. Most participants believe that current BCI technology cannot decode inner thoughts, although they acknowledge the potential for future advancements. Key technical challenges, such as signal quality and reliance on background information, are highlighted.

Conclusion: We summarize the interpretations, feasibility, and limitations of BMR and introduce a distinction between "strong BMR" and "weak BMR" to clarify their technical and ethical implications. Based on our analysis, we argue that current BMR does not pose unique ethical challenges compared with other forms of mind reading, and therefore does not yet justify the establishment of a distinct right to mental privacy.

Annotation

The finding: This qualitative study used semi structured interviews with 20 Chinese experts in BCI and neuroscience, then applied reflexive thematic analysis to characterize how experts interpret “mind reading,” how feasible they believe it is, and what technical limits currently constrain it.

Most experts held that present day BCIs cannot decode “inner thoughts,” and emphasized barriers such as inadequate signal quality and heavy dependence on background or contextual information.

The authors argue that, given these limitations, current BCI based mind reading does not yet present unique ethical challenges compared with other forms of mind reading and therefore does not currently justify a distinct right to mental privacy.

They propose a clarifying distinction between “weak BMR” (limited decoding of simpler mental states) and “strong BMR” (mind reading that could bypass user control), and outline why passive BCIs are the most plausible pathway to strong BMR.

Strength and weaknesses: Strengths include an empirically grounded contribution to a debate that is often driven by philosophical argument and media hype, plus a useful conceptual framework (strong vs weak BMR) that ties ethical concern to concrete technical features like user control and passive signal acquisition.

The paper also explicitly integrates technical considerations (signal quality, background information, user cooperation) into the ethical analysis, which helps readers avoid category errors about what current systems can actually do.

Limitations include a small sample (n=20) and reliance on qualitative interviews without accompanying quantitative comparisons, which constrains generalizability and makes it hard to estimate expert consensus across subfields.

The authors also note that their interviews focused on general BCI ethics rather than mind reading technologies specifically, which may limit the depth of detail on BMR specific ethical risks. Finally, the work is primarily descriptive and does not resolve the normative question of whether new rights are needed, even if it usefully narrows when that debate is most warranted.

Relevance: For consultation liaison psychiatrists, the clinical relevance is indirect but increasingly practical. C L psychiatrists are frequently asked to evaluate privacy, coercion, decisional capacity, and the meaning of sensitive data streams in medically complex settings, and neurotechnology is moving toward clinical and consumer deployment that could generate exactly those dilemmas. This paper offers a reality check that current BCI mind reading is limited and often requires cooperation and context, which can temper alarmist assumptions in bedside ethics discussions, policy conversations, and interdisciplinary consultation.

The strong versus weak BMR distinction is especially useful for C L psychiatrists advising hospitals on when a technology crosses from “just another sensitive data source” to a qualitatively higher risk domain because it may bypass user control, a shift the authors link to passive BCIs and loss of conscious control.

Organ Donation After Medical Aid in Dying: An Ethical Overview
David Rodríguez-Arias, María Victoria Martínez-López, Luis Espericueta, Gonzalo Díaz-Cobacho, Jed Adam Gross, Janet Delgado

Abstract:

Organ Donation after Medical Aid in Dying (OD-MAiD) is currently practised in four countries: Belgium, Canada, the Netherlands, and Spain. While OD-MAiD shares some similarities with MAiD (absent the possibility of organ donation) and with standard organ donation protocols, the combination of OD and MAiD involves unique circumstances that present novel ethical challenges. These challenges revolve around donors' consent and protection, the dead donor rule, and organ allocation. This paper explores these moral challenges and proposes strategies to ensure ethical safeguards in the context of OD-MAiD. An underlying question is whether OD-MAiD, if permitted, should follow the ethical guidelines of living donation or deceased donation, as these two practices commonly operate under distinct moral paradigms. While the living donation paradigm is centred on the protection of donors' interests and emphasises individual choice by allowing donors to decide who receives their organs, the deceased donation framework places more emphasis on enabling recipients to benefit from transplant, and organ allocation is typically based on impartiality. OD-MAiD also raises ethical concerns about how the possibility of donation could influence a patient's decision to seek euthanasia and/or interfere with optimal end-of-life care. Proposing organ donation to individuals considering MAiD could conceivably create pressure to proceed with euthanasia, either to realise a social good or to satisfy the needs of loved ones (if a family member requires an organ). This may undermine the patient's autonomy or well-being at the end of life.

Annotation

The finding: This article provides an ethical overview of organ donation after medical aid in dying (OD-MAID), arguing that OD-MAID raises distinct ethical challenges beyond standard living or deceased donation because donation decisions and end-of-life decisions can become tightly coupled. The authors map key pressure points, including risks that the option to donate could influence a patient’s MAID choice (or be experienced as a burden), the need for strict separation of roles and robust consent processes, and unresolved controversies about death determination and the “dead donor rule” in protocols designed to optimize organ viability.

Strength and weaknesses: A major strength is the article’s structured synthesis of the main ethical domains (autonomy and voluntariness, procedural safeguards, death determination, organ allocation, and public trust) and its practical orientation toward policy and protocol design rather than abstract principle-only argument.

Limitations are that it is a narrative ethics overview rather than empirical outcomes research, so it cannot quantify how often coercion, decisional conflict, or downstream harms occur, and many recommendations necessarily rest on contestable premises about MAID’s moral status and jurisdiction-specific legal frameworks, which limits portability across settings.

Relevance: For consultation-liaison psychiatrists, OD-MAID is a high-stakes intersection of capacity, voluntariness, depressive symptoms, family dynamics, and institutional risk, all of which are common C-L consultation targets. The paper is particularly relevant where C-L teams are asked to assess decision-making capacity, evaluate coercion or undue influence, help teams manage value conflict and moral distress, and advise on safeguards that protect patient autonomy while maintaining public trust in both MAID and organ donation systems.

Should We Use Behavioural Predictions in Organ Allocation?
Max Drezga-Kleiminger, Dominic Wilkinson, Thomas Douglas, Joanna Demaree-Cotton, Julian Koplin, Julian Savulescu

Abstract:

Medical predictions, for example, concerning a patient's likelihood of survival, can be used to efficiently allocate scarce resources. Predictions of patient behavior can also be used-for example, patients on the liver transplant waiting list could receive lower priority based on a high likelihood of non-adherence to their immunosuppressant medication regimen or of drinking excessively. But is this ethically acceptable? In this paper, we will explore arguments for and against behavioural predictions, before providing novel empirical evidence on this question. Firstly, we note that including behavioural predictions would lead to improved transplant outcomes. Fairness could also require prioritising those predicted to engage in healthier behaviours: consistent with using behavioural predictions in other contexts such as psychiatry and substance misuse. Conversely, behavioural predictions may be judged too inaccurate or discriminatory, or it may be thought unfair to deprioritise based on future behaviour. In part two, we performed an online survey of 172 UK adults. When presented with possible factors relevant to liver allocation, most thought predictions of higher medication adherence (78.6%) and lower future alcohol use (76.5%) should be used but not predictions of lower future criminality (24.7%) and higher societal contribution (21.2%). Randomising participants into two groups, 69.8% of participants found deprioritising a patient based on their predicted medication adherence acceptable (91.9% found a nonbehavioural prediction acceptable). We did not identify an ethically relevant difference between behavioural predictions and other medical predictions already used in organ allocation. Our sample of participants also appeared to support behavioural predictions in this context.

Annotation

The finding: This paper argues that incorporating behavioural predictions into liver transplant allocation (for example, predicted medication adherence or future alcohol use) can be ethically justified when the point is to improve predicted medical outcomes, and that there is no clear morally relevant distinction between these behavioural predictions and the “natural” medical predictions already used in allocation. In a survey of 172 UK adults, most respondents endorsed using predictions of higher medication adherence (78.6%) and lower future alcohol use (76.5%), but not predicted lower criminality (24.7%) or higher “societal contribution” (21.2%).

When participants were randomized to a “natural” versus “behavioural” scenario, 69.8% found deprioritizing a candidate based on predicted medication nonadherence acceptable, compared with 91.9% for deprioritizing based on an immune-system-based rejection prediction (suggesting greater comfort with non-behavioural predictors, even when both are framed as outcome-relevant).

Strength and weaknesses: A major strength is the integrated structure: a normative analysis that explicitly tests the “is there a morally relevant difference?” question, paired with empirical public-attitudes data (including a randomized comparison of behavioural versus non-behavioural framing).

The survey also usefully distinguishes behaviours plausibly linked to medical benefit (adherence, alcohol use) from factors that look more like social worth judgments (criminality, societal contribution), which many respondents rejected.

Key limitations are typical for a pilot survey: it is a modest, online convenience sample (N = 172) that is predominantly White and relatively highly educated, and responses are to simplified hypotheticals rather than real-world tradeoffs among multiple allocation criteria.

The paper also does not empirically validate the accuracy, calibration, or bias-properties of behavioural prediction tools themselves, which is central to the real ethical risk (especially disparate impact).

Relevance: For consultation-liaison psychiatrists, this is directly relevant to transplant psychiatry and to the broader ethics of using behavioural risk prediction in medicine. The paper explicitly links behavioural prediction in organ allocation to parallel practices in psychiatry and substance use care, which makes it a useful conceptual bridge for C-L clinicians who are asked to weigh in on adherence, relapse risk, or psychosocial factors in high-stakes medical decisions.

Practically, the findings help frame how to discuss “behavioural” criteria with teams and families: the public appears more supportive when behavioural predictions are tethered to medical benefit (adherence, alcohol use) and less supportive when they drift into social worth (criminality, societal contribution). This supports a C-L role in (1) clarifying which psychosocial variables are ethically defensible to consider, (2) anticipating stigma and structural inequity concerns, and (3) scrutinizing any AI-enabled risk tools for bias and legitimacy before they influence allocation.