Neurotechnologies such as brain-computer interfaces (BCIs) and deep brain stimulation (DBS) raise distinctive concerns about human agency. A recent proposal by Schönau et al. identifies four dimensions along which neurotechnologies may threaten agency: responsibility, privacy, authenticity, and trust. The Schönau et al. framework valuably maps how neurotechnologies affect users and has inspired qualitative assessment tools. Yet although it acknowledges interconnections among the four dimensions, it does not explain what structurally unifies them. This represents a key gap in the view. Assessment instruments modeled on the framework, such as their Qualitative Agentive Competency Tool (Q-ACT), evaluate each dimension independently, risking fragmented assessments that miss the integrated nature of agential harm. I argue that agential self-trust unifies these four dimensions. Each tracks a distinct way that neurotechnologies can erode the self-trust constitutive of planning agency: confidence in one's control over action (responsibility), in the boundaries of one's deliberative life (privacy), in one's psychological continuity (authenticity), and in one's sensory and evaluative capacities (trust). This carries concrete policy implications. Current informed consent procedures for neurotechnology trials enumerate risks along separate dimensions without flagging the cumulative threat to a user's capacity for agential self-trust. Likewise, assessment instruments should include integrative measures that track agential self-trust across domains. As neurotechnology governance develops at the international level, a unified account of agential harm within this dimensional framework can guide both consent design and longitudinal monitoring.
For children with severe physical and cognitive disabilities (e.g. iatrogenic neurologic injury, congenital myelopathy, or quadriplegic cerebral palsy), there are limited therapeutic options. Scientific and clinical developments in implantable brain-computer interface (BCI) technology are in clinical trials in adults. The ethical issues around implantable BCI research in adults have recently been examined, but to date, only limited literature is available on the ethical issues that are attendant with implantable pediatric BCI research. Here, we summarize the ethical issues, focusing on (1) whether invasive BCI research should proceed in children, (2) regulatory considerations, (3) study design considerations, (4) pediatric recipient selection for invasive BCI trials, (5) special problems regarding informed consent in this context, and (6) related psychosocial and public perception considerations. We conclude with specific recommendations regarding ethically informed design of invasive pediatric BCI trials.
Bioprediction uses biomedical data, biomarkers, and algorithmic tools to forecast future physiological states or impairments. As predictive technologies increasingly enable individuals to anticipate loss of cognitive capacity before it occurs, they raise a fundamental ethical question: how should moral responsibility be assessed when impairment becomes foreseeable in advance of its onset? We argue that biopredictive technologies can alter both the epistemic and control conditions that ground moral responsibility. In some cases, this generates obligations to respond to predictive warnings of cognitive decline; in more limited circumstances, it may also ground obligations to acquire and use biopredictive technologies where these represent a proportionate and least restrictive means of risk reduction. We further argue that emerging individual responsibilities generate corresponding obligations for states. Where access to biopredictive technologies is necessary to meet morally significant expectations of risk management, fairness, and legitimacy require attention to affordability, accessibility, and institutional support.
Biomarker-based technologies for predicting age of onset are currently being developed for carriers and individuals at 50% risk of autosomal dominant neurodegenerative diseases such as Huntington's disease, frontotemporal dementia and spinocerebellar ataxias. Qualitative interview studies indicate that carriers and individuals at 50% risk expect that onset predictions would be valuable for life planning but would also impact mental health. Ethically responsible implementation of biomarker-based onset prediction in research and clinical settings requires maximizing the utility and minimizing negative psychosocial effects. We present a framework specifying features that influence the value and impact of onset predictions, grouped into four domains: features of the disease, the onset-predictive test, the individual, and the context. The framework is intended to support ethicists, researchers, developers, and healthcare professionals in anticipating and addressing the ethical implications of current and emerging onset prediction technologies for individuals at risk of genetic neurodegenerative disease.
Noninvasive neuromodulation techniques such as transcranial magnetic stimulation (TMS) and transcranial direct current stimulation (tDCS) are well-established neuroscience research tools. Because both rely on scalp contact, some have speculated that they may not work effectively across all hair types and textures. The present study investigated the presence of phenotypic bias in these technologies by interviewing TMS (n = 22) and tDCS (n = 16) researchers about their experiences administering them. The majority (71.1%) reported encountering difficulties administering neurostimulation due to research subjects' physical features, most frequently hairstyles common among Black research subjects. Half agreed that it is more difficult to administer TMS or tDCS to individuals from certain racial and ethnic backgrounds. Among participants who did not report difficulty, advance communication and screening was the most common strategy for avoiding challenges. Our findings provide empirical support for concerns about phenotypic bias in noninvasive neuromodulation; future efforts should identify and address factors underlying these difficulties.
Since its introduction, electroconvulsive therapy (ECT) has been controversial. Recently, a report by the World Health Organization called for an ECT ban for children and adolescents. This has added a new layer to consider in the debate about the ethical use of ECT. To better understand ethical concerns and attitudes toward the use of ECT in this population, and to assess whether a ban might be justified, we conducted a comprehensive literature review. Close to 75% of our analyzed papers reported agreement about ECT's effectiveness for treating a wide range of adolescent mental disorders. Almost half of the papers reported memory side effects as an ethical concern. In addition, the lack of rigorous side-effect reporting was found to be a key barrier to the ethical use of ECT in this population. We hope this review will advance our understanding of ongoing ethical concerns and barriers regarding use of ECT in adolescents.
Closed-loop neurotechnologies bring great promise for treating neurological and psychiatric disorders. However, the use of artificial intelligence (AI) in their application raises ethical concerns, since AI-driven closed-loop devices may cause unforeseen mental interference that might, absent consent, infringe the user's mental rights. Whether such worries are warranted, however, may depend on whether closed-loop neurotechnologies qualify as moral agents, and on whether they are distinct from the moral agent on whose brain they act. If they are not moral agents, or are not separate moral agents, they will arguably be incapable of infringing the user's mental rights. In this article, we explore different possible agential relationships between the human user and closed-loop neurotechnologies and consider the implications for the protection that our mental rights provide.
AI chatbots are rapidly being considered and increasingly used for mental health support by professionals, patients, and others. As the utility, advantages, as well ask ethical risks of AI chatbots vary per stakeholder, we sought to explore attitudes toward the applications, benefits, concerns of, and requirements for AI chatbots use for mental health support among different key stakeholders. We conducted a multi-stakeholder qualitative survey with three groups: mental health patients (n=40), potential users (n=48), and mental health professionals (n=32). The different groups listed very similar benefits and concerns of AI chatbots use. The survey revealed unique themes such as the value of maintaining human-human interaction in therapy. Autonomy, often highlighted as a core ethical principle in AI, was not strongly emphasized by respondents. Future research should examine how users engage with AI chatbots in practice, and how guidance should be developed in response to this rapidly changing technology.
