ERT

'Training Humans to Detect Children's Lies Through their Facial Expressions' by Alison O'Connor, Kaila Bruer, Jennifer Gongola, Thomas D Lyon and Angela D Evans in Applied Cognitive Psychology (in press) comments 

 The accurate detection of children’s truthful and dishonest reports is essential as children can serve as important providers of information. Research using automated facial coding and machine learning found that children who were asked to lie about an event were more likely to look surprised when hearing the first question during an interview about said event. The present studies explored if humans can be trained to look for surprised expressions to detect children’s deception. Participants made lie-detection judgments after seeing children’s expressions in very brief clips. In Study 1, we compared performance across a training condition and control condition, and in Study 2 we modified the training. With training, adults could detect children’s lies at above chance levels by viewing their facial expression. Detection accuracy was further improved with modified training (Study 2), but participants held a consistent lie bias. Challenges with using facial expressions to detect deceit are discussed.

Neurotech

'Neurotechnologies and human rights: restating and reaffirming the multi-layered protection of the person' by Christoph Bublitz in (2024) 28(5) The International Journal of Human Rights 782 comments

Advances in neuroscience and neurotechnology and the significant powers they confer over minds and bodies of persons have given rise to grave concerns and caught the attention of lawmakers, ethicists, international organisations, and the human rights community. Following recent reports and statements by the OECD, the International Bioethics Committee of UNESCO (IBC), the Interamerican Juridical Committee of the Organization of American States (IJC), the Council of Europe and the European Parliament which highlight a range of threats to human rights posed by neurotechnologies, the UN Human Rights Council requested its Advisory Committee to prepare a comprehensive study on the topic. The IBC recommends, among other measures, to further develop the interpretation of existing rights and ‘the adaptation of existing human rights instruments and the proclamation of new human rights’, e.g. in a new ‘Universal Declaration on Human Rights and Neurotechnology’. Heeding the recommendation, UNESCO has started to develop the first global standard-setting document on the ethics of neurotechnologies, and the idea about a subsequent binding international treaty is in the air. 

Current interest in neurotechnologies arises from their potential to affect central characteristics of persons such as altering their thoughts and feelings, their bodily and mental capacities, their memories or personalities, as well as the technology’s potential to reveal and expose supposedly private aspects of the human mind. The technology develops fast, increasingly implements machine learning methods (Artificial Intelligence, AI), and is driven by start-ups with significant investments by venture capital. More and more products become mature for market release as medical or consumer devices. The attention of the UN, human rights and regulatory agencies is thus timely and warranted. A key topic in the aforementioned reports, crucial for policy and regulatory activities, is the worry that current human rights law might be insufficient to protect persons against many conceivable attacks with the help of neurotechnologies. In this vein, some scholars and neuroscientists have suggested that human rights law has substantive gaps that require closing by the introduction of novel human rights, so called ‘neurorights’, which has become a summary term for novel but vague rights that apply to neurotechnologies. The call for novel rights is grounded in a narrative of deficiency of existing rights. But whether this narrative is persuasive and whether current law indeed leaves substantial gaps that require closing is an open question. Although it is logically prior to the call for novel rights, it has not received thorough examination from neurorights advocates or legal scholars. Providing an analysis of how established human rights relate to neurotechnological challenges is the aim of the following. 

Contrary to the narrative of the deficiency of established rights, the analysis will show that reasonably constructed, established rights provide a nuanced multi-layered protection of the person, including her mind, against virtually all conceivable threats by neurotechnologies. More concretely, it will be argued that all worrisome uses of neurotechnologies either interfere with the rights to bodily and mental integrity or the right to privacy. In addition, severe interferences may affect the right to freedom of thought or human dignity. As not all of these rights have been fully examined by human rights scholarship, tentative definitions will be proposed. that render them, especially the nascent right to mental integrity and the long-established yet practically irrelevant freedom of thought, applicable to neurotechnological challenges. Moreover, the paper will offer four readings of human dignity that may inform further interpretations of specific rights and revolve around the idea of protecting individuals as persons and subjects, respecting their subjectivity while rejecting their objectification. In sum, this analysis restates existing rights and reaffirms their applicability to neurotechnologies; it demonstrates the adaptability of established rights to novel circumstances without contravening the text of provisions or the spirit of instruments; on the contrary, it makes explicit the multi-layered protection of the person that is deeply entrenched in core instruments. Accordingly, considerable gaps in the protection of the person at the level of generality of human rights that require closing by novel rights are not evident. Instead, novel technologies invite and require the actualisation and specification of existing guarantees. 

Whether these proposed constructions of rights will stand the test of time and be accepted by courts of course remains to be seen. Real case material does not exist yet, and in its absence, only an abstract anticipatory analysis of the scope and limits of rights is possible. It must leave out the many context-specific considerations that may arise in the many conceivable cases in which neurotechnologies may affect persons. The argument in the following lies on a different plane, it argues that defining features of the modus operandi of neurotechnologies makes any worrisome use of them fall under the ambit of established rights. The proposed constructions of rights may guide courts and policymakers, inspire rightholders to invoke them in legal proceedings, and redirect the current discourse about novel rights by debunking its premise of the deficiency of existing rights.

Surveillance

'Big brother: the effects of surveillance on fundamental aspects of social vision' by Kiley Seymour, Jarrod McNicoll and Roger Koenig-Robert in (2024) 1 Neuroscience of Consciousness argues 

Despite the dramatic rise of surveillance in our societies, only limited research has examined its effects on humans. While most research has focused on voluntary behaviour, no study has examined the effects of surveillance on more fundamental and automatic aspects of human perceptual awareness and cognition. Here, we show that being watched on CCTV markedly impacts a hardwired and involuntary function of human sensory perception—the ability to consciously detect faces. Using the method of continuous flash suppression (CFS), we show that when people are surveilled (N = 24), they are quicker than controls (N = 30) to detect faces. An independent control experiment (N = 42) ruled out an explanation based on demand characteristics and social desirability biases. These findings show that being watched impacts not only consciously controlled behaviours but also unconscious, involuntary visual processing. Our results have implications concerning the impacts of surveillance on basic human cognition as well as public mental health. .. 

In recent years, we have seen an exponential increase in human surveillance. We now live in a world with closed-circuit television (CCTV) in public spaces, trackable mobile devices, and the monitoring of our activities through artificially intelligent technology and the ‘Internet of Things’ (the interconnected system of our devices and sensors collecting and sharing data through the internet). Data on what we do, what we say, and where we go can be monitored and made available to third parties (Zuboff 2015, Cecez-Kecmanovic 2019). With the advent of emerging neurotechnology, even our mental privacy is at risk (Farahany 2023). Despite this proliferation of surveillance technology, there is limited research on its effects on human psychology, including fundamental capacities like the basic perceptual processing of our sensory environment. 

Literature available on the topic of human surveillance and being watched suggests that it elicits changes in overt behaviour. For instance, a large body of evidence on ‘audience effects’ suggests people act in a more prosocial manner when they believe they are being watched. When people think their behaviour is monitored, they are more giving (Hoffman et al. 1996, Haley and Fessler 2005, Pfeiffer and Nowak 2006, Rigdon et al. 2009, Powell et al. 2012, Nettle et al. 2013, Bateson et al. 2015), more likely to share (Baillon et al. 2013, Oda et al. 2015), and less likely to steal, cheat, litter, or direct their gaze to provocative images (Tourangeau and Yan 2007, Zhong et al. 2010, Risko and Kingstone 2011, Francey et al. 2012, Nettle et al. 2012, Nasiopoulos et al. 2015). It is argued that these behavioural changes act to uphold the reputation of the individual and protect from negative social consequences (Izuma 2012, Nettle et al. 2013, Conty et al. 2016). 

In addition to the changes in social behaviour, a feeling of being watched commonly invokes discomfort in people (Panagopoulos and Van Der Linden 2017) and increases vigilance, self-consciousness, and the fight-or-flight response (e.g. an increase in heart rate and skin conductance) (Kleinke and Pohlen 1971, Nichols and Champness 1971, Gale et al. 1975, Putz 1975, Reddy 2003, Conty et al. 2010b, Helminen et al. 2011, Baltazar et al. 2014). It has also been shown that surveillance in the workplace induces negative effects on productivity (Gagné and Deci 2005), likely due to impacts on attention and working memory (Senju and Hasegawa 2005, Conty et al. 2010a, Risko and Kingstone 2011, Wang and Apperly 2017, Colombatto et al. 2019). Interestingly, it seems to be an implied social presence rather than a true presence of the observer that is important here, with simple photos of watching eyes or a mere belief that someone is watching eliciting the behavioural changes (Putz 1975, Haley and Fessler 2005, Bateson et al. 2006, Burnham and Hare 2007, Rigdon et al. 2009, van Rompay et al. 2009, Risko and Kingstone 2011, Lawson 2015, Nasiopoulos et al. 2015, Colombatto et al. 2019). 

While the effects of surveillance on social behaviour are well-documented, it is unclear how being watched impacts more fundamental capacities not subject to explicit, overt, and conscious control of the individual. For instance, being able to rapidly detect when someone or something is looking at you is a profound and hardwired human faculty requiring specialized neural mechanisms that operate largely outside of conscious control (Brothers 1990, Perrett and Emery 1994, Baron-Cohen 1995, Emery 2000, Calder et al. 2007, Hietanen et al. 2008, Senju and Johnson 2009, Bayliss et al. 2011, Burra et al. 2013, Carlin and Calder 2013). In fact, this heightened sensitivity to another’s gaze is thought to underlie a feeling of being watched that can be experienced in the absence of any surveillance and commonly reported in the population (Freeman et al. 2005, Taylor et al. 2009, Bebbington et al. 2013, Harper and Timmons 2021). Given the adaptive significance, we hypothesize that these mechanisms are further engaged when one knows they are being watched. Indeed, evidence from clinical research suggests that patients with schizophrenia who experience persecutory delusions (i.e. erroneous beliefs about being watched) show increased perceptual sensitivity to the self-directed gaze of others (Rosse et al. 1994, Hooker and Park 2005, Tso et al. 2012, Langdon et al. 2017). 

In the current study, we test whether being watched influences perceptual processing of the sensory environment, namely the processing of eye gaze. Specifically, we ask whether being monitored makes the visual system more sensitive to this essential visual and social cue. Using a technique known as breaking continuous flash suppression or b-CFS (Tsuchiya and Koch 2005), we temporarily suppressed photographs of faces from visual awareness. The time the face takes to break through the suppressive mask and become visible to the participant is typically treated as an index of its salience. In previous experiments using b-CFS, it has been shown that the visual system prioritizes the detection of faces with direct gaze over faces with averted gaze, suggesting visual cues used to discriminate eye gaze direction are preconsciously processed by the visual system (Stein et al. 2011, Yokoyama et al. 2013, Seymour et al. 2016). In the current study, we examined whether being watched influenced the speed at which these gaze signals reach conscious awareness by means of a detection task (i.e. stimulus on the left or right). We hypothesized that if being surveilled facilitates basic sensory processing of eye gaze, then participants who had evidence of being monitored during the task (i.e. experiencing the presence of CCTV) would detect self-directed gaze signals faster than participants who did not.

Emotion

'Physiognomic Artificial Intelligence' by Luke Stark and Jevon Hutson in (2022) 32 Fordham Intellectual Property, Media and Entertainment Law Journal 922 comments 

The reanimation of the pseudosciences of physiognomy and phrenology at scale through computer vision and machine learning is a matter of urgent concern. This Article—which contributes to critical data studies, consumer protection law, biometric privacy law, and antidiscrimination law—endeavors to conceptualize and problematize physiognomic artificial intelligence (“AI”) and offer policy recommendations for state and federal lawmakers to forestall its proliferation. 

Physiognomic AI, as this Article contends, is the practice of using computer software and related systems to infer or create hierarchies of an individual’s body composition, protected class status, perceived character, capabilities, and future social outcomes based on their physical or behavioral characteristics. Physiognomic and phrenological logics are intrinsic to the technical mechanism of computer vision applied to humans. This Article observes how computer vision is a central vector for physiognomic AI technologies and unpacks how computer vision reanimates physiognomy in conception, form, and practice and the dangers this trend presents for civil liberties. 

This Article thus argues for legislative action to forestall and roll back the proliferation of physiognomic AI. To that end, it considers a potential menu of safeguards and limitations to significantly limit the deployment of physiognomic AI systems, which hopefully can be used to strengthen local, state, and federal legislation. This Article foregrounds its policy discussion by proposing the abolition of physiognomic AI. From there, it posits regimes of U.S. consumer protection law, biometric privacy law, and civil rights law as vehicles for rejecting physiognomy’s digital renaissance in AI. Specifically, it contends that physiognomic AI should be categorically rejected as oppressive and unjust. Second, it argues that lawmakers should declare physiognomic AI unfair and deceptive per se. Third, it proposes that lawmakers should enact or expand biometric privacy laws to prohibit physiognomic AI. Fourth, it recommends that lawmakers should prohibit physiognomic AI in places of public accommodation. It also observes the paucity of procedural and managerial regimes of fairness, accountability, and transparency in addressing physiognomic AI and attend to potential counterarguments in support of physiognomic AI.

The robust and important 'Neurorights: The Land of Speculative Ethics and Alarming Claims?' by Frederic Gilbert and Ingrid Russo in (2024) 15(2) AJOB Neuroscience 113 comments 

 The intersection of AI and neurotechnology has resulted in an increasing number of medical and non-medical applications and has sparked debate over the need for new human rights, or “neurorights,” to better protect users. In his article, Bublitz critically examines the prospect of an international instrument regarding Neurotechnologies and Human Rights. In evaluating the feasibility of establishing new human rights, Bublitz argues in favor of advancing the law without introducing novel rights (Bublitz 2024). He acknowledges the criticality of protecting certain fundamental aspects of the mind—specifically, the unconditionally protected core of freedom of thought and opinion—alongside qualified rights to mental integrity and privacy which protect against less severe neurotechnological interferences (Bublitz 2024). In this commentary, we build upon Bublitz’s position by examining the calls for new human rights based on assertions that the mind requires safeguarding from invasive ‘reading’ technologies (Bublitz 2024). 

Let us look at the ‘reading’ terminology used in these assertions. First, we need to delve into the veracity of terms like “brain-reading” and “mind-reading” in the context of neurotechnological advancements to discern whether the claims are underpinned by evidence or hype. In recent years, there has been a surge in news media reports discussing the potential of AI applications to decode brain activity for mind-reading purposes. The portrayal of AI mind-reading capabilities is both remarkable and concerning. Recent headlines, such as “The brain is the final frontier of our privacy, and AI is about to breach it” (Yahoo News), “Mind-reading technologies have arrived” (VOX), “AI makes non-invasive mind-reading possible by turning thoughts into text” (The Guardian), “This ‘mind-reading’ AI system can recreate what your brain is seeing” (Euronews), “AI-Powered ‘Thought Decoders’ Won’t Just Read Your Mind—They’ll Change It” (Wired), are so commonplace that one feels ‘AI ability to read the mind’ is mainstream reality. 

However, given that news media also often depict brain-computer interfaces (BCIs) in an unjustifiably overall positive and sensationalist tone, a degree of skepticism arises regarding the claims that AI can access and decrypt hidden aspects of the mind (Gilbert et al. 2019; Pham and Gilbert 2019). 

Interestingly, the claims about AI-enabled mind-reading find resonance even within the most respected and influential institutions. For instance, the International Bioethics Committee of UNESCO’s report on ‘The Risks and Challenges of Neurotechnologies for Human Rights’ underscores the multifaceted impacts of combining AI and neurotechnologies capable of ‘reading’ and ‘writing’ brain activity. Furthermore, academic journals contribute to this discourse, with titles like “Mind-reading devices are revealing the brain’s secrets” (Nature) and “Artificial intelligence is learning to read your mind—and display what it sees” (Science). 

We conducted a scoping review of 1017 academic articles to gain insights into the current state of the art and examine assertions made by academics (Gilbert and Russo under review). Our analysis revealed that up to 91% of the examined articles suggest the possibility of mind reading through brain reading (Figure 1). Overall, we observed an increase in the number of articles connecting brain reading and mind reading by year (Figure 2), along with discussion that mind-reading will be possible in the future (Figure 3). Ethical issues discussed frequently include mental privacy, mental freedom, and personhood.

Emotion Recognition

'The unbearable (technical) unreliability of automated facial emotion recognition' by Federico Cabitza, Andrea Campagner and Martina Mattioli in (2022) 9(2) Big Data and Society comments 

Emotion recognition, and in particular facial emotion recognition (FER), is among the most controversial applications of machine learning, not least because of its ethical implications for human subjects. In this article, we address the controversial conjecture that machines can read emotions from our facial expressions by asking whether this task can be performed reliably. This means, rather than considering the potential harms or scientific soundness of facial emotion recognition systems, focusing on the reliability of the ground truths used to develop emotion recognition systems, assessing how well different human observers agree on the emotions they detect in subjects’ faces. Additionally, we discuss the extent to which sharing context can help observers agree on the emotions they perceive on subjects’ faces. Briefly, we demonstrate that when large and heterogeneous samples of observers are involved, the task of emotion detection from static images crumbles into inconsistency. We thus reveal that any endeavour to understand human behaviour from large sets of labelled patterns is over-ambitious, even if it were technically feasible. We conclude that we cannot speak of actual accuracy for facial emotion recognition systems for any practical purposes. ... 

Emotional artificial intelligence (AI) (McStay, 2020) is an expression that encompasses all computational systems that leverage ‘affective computing and AI techniques to sense, learn about and interact with human emotional life’. Within the emotional AI domain (but even more broadly, within the entire field of AI based on machine learning (ML) techniques), acial emotion recognition (FER), which denotes applications that attempt to infer the emotions experienced by a person from their facial expression (Paiva-Silva et al., 2016; McStay, 2020; Barrett et al., 2019), is one of the most controversial (Ghotbi et al., 2021) and debated (Stark and Hoey, 2021) applications. 

In fact, ‘turning the human face into another object for measurement and categorization by automated processes controlled by powerful companies and governments touches the right to human dignity’ and ‘the ability to extract […physiological and psychological characteristics such as ethnic origin, emotion and wellbeing…] from an image and the fact that a photograph can be taken from some distance without the knowledge of the data subject demonstrates the level of data protection issues which can arise from such technologies’. On the other hand, opinions diverge among the specialist literature. Some authors highlight the accurate performance of FER applications and their potential benefits in a variety of fields; for instance, customer satisfaction (Bouzakraoui et al., 2019), car driver safety (Zepf et al., 2020), or the diagnosis of behavioural disorders (Paiva-Silva et al., 2016; Jiang et al., 2019). Others have raised concerns regarding the potentially harmful uses in sectors such as human resource (HR) selection (Mantello et al., 2021; Bucher, 2022), airport safety controls (Jay, 2017), and mass surveillance settings (Mozur, 2020). In addition, the scientific basis of FER applications has been called into question, either by equating their assumptions with pseudo-scientific theories, such as phrenology or physiognomy (Stark and Hutson, Forthcoming), or by questioning the validity of the reference psychological theories (Barrett et al., 2019), which assume the universality of emotion expressions through facial expressions (Elfenbein and Ambady, 2002). Lastly, others have noted that the use of proxy data (such as still and posed images) to infer emotions should be supported by other contextual information (McStay and Urquhart, 2019), especially if the output of the FER systems is used to make sensitive decisions, so as to avoid misinterpretation of the broader context. According to Stark and Hoey (2021) ‘normative judgements can emerge from conceptual assumptions, themselves grounded in a particular interpretation of empirical data or the choice of what data is serving as a proxy for emotive expression’. 

From a technical point of view, FER is a measurement procedure (Mari, 2003) in which the emotions conveyed in facial expressions are probabilistically gauged to detect the dominant one or a collection of prevalent emotions. As a result, FER can be related to the concepts of validity and reliability. A recognition system is valid if it recognizes what it is designed to recognize (i.e. basic emotions); it is reliable if the outcome of its recognition is consistent when applied to the same objects (i.e. a subject’s expression). However, when FER is achieved by means of a classification system based on ML techniques, its reliability cannot (and should not) be separated from the reliability of its ground truth, i.e. training and test datasets (Cabitza et al., 2019). In this scenario, reliability is defined as the extent to which the categorical data from which the system is expected to develop its statistical model are generated from ‘precise measurements’, i.e. human ‘recognitions’ exhibiting an acceptable agreement. This is because, by definition, no classification model can outperform the quality of the human reference (Cabitza et al., 2020b). 

In this study, we will not contribute to the vast (and heated) debate still currently going on about the validity of automatic FER systems (Franzoni et al., 2019; Feldman Barrett, 2021; Stark and Hoey, 2021), that is, we do not address the classification task from the conceptual point of view (how to define emotions, if possible at all) nor merely from the technical point of view (how to recognize emotions, whatever they are). For the sake of argument, we assume that the main psychological emotion models make perfect sense and we do not address how robust recognition algorithms are, how well they perform in external settings, and, most importantly, how useful they can be, i.e. whether they provide the benefits that their promoters envision and advocate. 

Instead, we focus on the reliability of their ground truth, which is not a secondary concern from a pragmatic standpoint (Cabitza et al., 2020a, 2020b). To that end, we conducted a survey of the major FER datasets concentrating on their reported reliability as well as a small user study by which we address three related research questions: Do existing FER ground truths have an adequate level of reliability? Are human observers in agreement regarding the emotions they sense in static facial expressions? Do they agree more when the context information is shared before interpreting the expressions? 

The first question is addressed in the ‘Related work and motivations’ section and the answer is in Table 3. The other questions are addressed by means of a user study described in the ‘User study: Methods’ section and whose results are reported in the ‘Results’ section. Finally, in the ‘Discussion’ section, we discuss these findings and their immediate implications, while in the ‘Conclusion’ section we interpret them within the bigger picture of FER reliability and relate them to implications for the use of automated FER systems in sensitive domains and critical human decision making.

'What an International Declaration on Neurotechnologies and Human Rights Could Look like: Ideas, Suggestions, Desiderata' by Jan Christoph Bublitz in (2024) 15(2) AJOB Neuroscience 96 comments 

 Ethical and legal worries arising from novel neurotechnological applications have reached the level of international human rights institutions and prompted ongoing deliberations about a new legal instrument that sets international standards for the development, regulation, and use of neurotechnologies. In a recent report on human rights implications of neurotechnologies, the International Bioethics Committee of UNESCO (IBC) considers the idea of a “governance framework set forth in a future UNESCO Universal Declaration on the Human Brain and Human Rights” or a “New Universal Declaration on Human Rights and Neurotechnology” (2021, at 184c). Other human rights agencies have been concerned with the matter, hosted hearings and commissioned reports (especially OECD  2019; see also Ienca  2021; OECD  2017; Sosa et al.  2022). The UN Human Rights Council (2022) mandated its Advisory Committee to prepare a comprehensive study on neurotechnologies and human rights. A novel international instrument will likely emerge from these debates (cf. UNESCO Docs. 216 EX/Dec.9 and EX/50). As the first global instrument specifically tailored to neurotechnologies, it will set the tone for further regulations at domestic, supranational, and international levels. Although some stakeholders have been consulted in previous proceeedings, the development has so far largely evaded the broader attention of the neuroscience, neurotech, and neuroethics communities. This is unfortunate, as academic input is vital to identify problems, frame debates and develop solutions, not least because international agencies lack subject matter expertise and have relied on a limited number of experts so far. The timing is critical. Once debates move to the political arena and intergovernmental negotiations, the room for academic and big picture debates narrows as matters tend to become increasingly technical and arguments tend to become interest-based. Accordingly, the time for impactful academic interventions is now. To facilitate it and to widen the perspective of current debates, this target article puts to discussion twenty-five considerations and desiderata for a future instrument. In particular, it wishes to transcend the confines of the debate about so called neurorights that dominates the current discourse (e.g., Borbón and Borbón  2021; Bublitz  2022c; Genser, Herrmann, and Yuste  2022; Ienca  2021; Ligthart et al.  2023; Rommelfanger, Pustilnik, and Salles 2022; Yuste et al. 2017; Zúñiga-Fajuri et al. 2021). Proceeding on the basis of existing rights, the following remains uncommitted as to whether novel rights are needed. This debate overshadows a broader and richer field of relevant questions, and it is time to turn to them. 

Setting the stage, the nature and the limits of a future instrument should be clarified. It will likely be a soft law instrument such as a recommendation by UNESCO or a resolution by the UN General Assembly. Such documents are not legally binding and lack enforcement mechanisms. Whether they qualify as law at all depends on legal theory’s perennial question about the nature of law and may be answered differently with respect to different types of documents (Andorno  2012; Shelton  2008). Suffice it to note here that such documents understand themselves as more than mere ethical statements because they demand compliance by signatory States without creating enforceable legal obligations. Theoretical matters aside, soft law instruments can be practically effective governance tools that draw attention to problems and set standards which are often observed by States and other stakeholders. They may, for instance, affect governmental research funding, decisions by ethics committees, or the regulatory conditions for market approval of devices. Soft law may also turn into hard law in several ways. It may provide guidance for courts in interpreting norms, rendering the content of rights more concrete and resolving normative conflicts. It may inform secondary soft law such as general comments by treaty bodies, and inspire further binding acts at domestic or international levels. Soft law’s greater flexibility is an advantageous feature in fast-moving fields without firm normative underpinnings such as neurotechnologies, and has therefore become the prime legal-regulatory tool for technology governance at both the international and the domestic level (Hagemann and Skees  2018; Marchant and Tournas 2019). At any rate, because of the often insurmountable political hurdles that binding treaties of international law face, especially in the current geopolitical climate, soft law instruments are the best form of international governance of neurotechnologies that is realistically attainable in the near future. 

The nature of an instrument shapes its content. In contrast to the abstract and elegantly worded Universal Declaration of Human Rights and the international covenants that followed it, soft law instruments allow for more aspirational goals and broader scopes but also for more concrete norms and standards. In addition, they are not only directed at States as the protagonists of international law but also at other stakeholders, notably private actors such as businesses that may threaten human rights, individuals whose rights may have been violated, but also other relevant parties such as engineers and developers of neurotechnologies. Moreover, given the aspiration of global applicability and the need for consensus in matters about which countries and cultures may reasonably disagree, instruments must allow for local adaptability, value pluralism, compromises, and gravitate toward smallest common denominators. These conditions are reflected in the texts of such documents, which are often replete with references to general values of the human rights systems, not always entirely coherent, and sometimes even intentionally vague at critical points. But despite and because of these weaknesses, soft law instruments can set norms and standards that are observed and steer the course of the future development of a field. The UNESCO Recommendation on Artificial Intelligence (AI), adopted in 2021, may serve as a model for a future neurotech instrument. It contains recommendations at different levels of abstraction, from broad values over principles to actionable policy options. Although not free from textual weaknesses, the Recommendation provides some novel, concrete, and surprisingly far-reaching standards. 

It is further worth noting that international norms for the regulation of neurotechnologies already exist. Current debates sometimes evoke the impression that they develop in a legal vacuum, but this is a bit misleading. For instance, placing devices on markets is regulated by domestic and supranational device regulation, such as the EU Medical Device Regulation, which covers neurotechnologies for medical and some non-medical purposes (European Union  2017). It leaves neurodevices for non-medical neuroimaging outside of its scope, but this is not a gap but rather an intentional regulatory decisions. At the international human rights level, the Oviedo Convention on Human Rights and Biomedicine (1997), a legally binding international treaty signed by more than 30 States, seeks to safeguard the dignity and integrity of persons “with regard to the application of biology and medicine” (Council of Europe 1997, preamble). Likewise, the non-binding UNESCO Universal Declaration on Bioethics and Human Rights (2005) was adopted in view of the “rapid advances in science and their technological applications” (2005, preamble). Both instruments contain various norms about human rights and informed consent that apply to neurobiological interventions. The same is true for the Recommendation on Responsible Innovation in Neurotechnology (OECD  2019). This leads to the first desideratum: (i) A future instrument should cohere with existing instruments but not merely repeat them; it should neither contradict them without compelling reasons, nor address similar points by different terms, and should strive to go beyond them by suggesting more concrete norms or addressing substantially different aspects. 

The following presents further desiderata and considerations for a future instrument. It proceeds from the general to the particular, from meta-considerations to concrete rights and technical suggestions, and at least partially attempts to deduce the latter from the former. The points are thus interwoven rather than distinct; they are sometimes couched in the idiosyncratic style of international documents and should not be understood as conclusive but as an invitation for criticism and additions.

Bioethics

'Equality-enhancing potential of novel forms of assisted gestation: Perspectives of reproductive rights advocates' by Elizabeth Chloe Romanis in (2023) Bioethics comments 

Novel forms of assisted gestation—uterus transplantation and artificial placentas—are highly anticipated in the ethico-legal literature for their capacity to enhance reproductive autonomy. There are also, however, significant challenges anticipated in the development of novel forms of assisted gestation. While there is a normative exploration of these challenges in the literature, there has not yet, to my knowledge, been empirical research undertaken to explore what reproductive rights organisations and advocates identify as potential benefits and challenges. This perspective is invaluable. These organisations/individuals have an awareness not only of the needs of individuals but also of the political landscape in which regulatory decisions are made and which individuals navigate when seeking reproductive assistance. In this study, data was generated from two semi-structured focus groups (n = 11). Reflective thematic analysis was used to examine the views raised by study participants in these focus groups. This paper explores two of the themes constructed in the data. First, the equality-enhancing potential of assisted gestation exploring the multifaceted ways in which assisted gestation has structural benefits for marginalised groups. Second, realising the equality-enhancing potential of assisted gestation explores the intersecting barriers to access to reproductive technologies and how they may impede the benefits of these technologies in practice. These results can enhance conceptual understanding of the importance of novel forms of assisted gestation and ensure that attention is paid to practical barriers in further normative research.  

 

For people unable to gestate, whether for biological, social or psychosocial reasons, assisted gestation affords them the opportunity to become biological parents. Assisted gestation has long existed in surrogacy. However, novel reproductive technologies that make possible different forms of assisted gestation are on the horizon and are highly anticipated for the specific experiences they could afford people who cannot gestate. This article reports results from an empirical research project seeking the views of reproductive rights advocates in Great Britain about the benefits and limitations of novel forms of assisted gestation. 

 

Uterus transplantation (UTx) presents the possibility of enabling those without a uterus of their own (whether they were born without it or have had a hysterectomy) to gestate—experiencing pregnancy and birth. Since the first report of a live birth from a successful UTx in Sweden in 2014, approximately one hundred transplants have been successfully performed worldwide. While the procedure is not yet widely available and is yet to be performed in the United Kingdom (at the time of writing), it is described as an ‘emerging therapy that is transitioning from an experimental phase to an established clinical practice’. To date, the transplant has only been performed on people with physiology assigned female at birth (AFAB). However, one team of research surgeons have concluded that despite being challenging ‘there is no overwhelming clinical argument against performing UTx’ on a person with physiology assigned male at birth. 

 

A more futuristic endeavour is the artificial placenta, capable of facilitating gestation outside of the body (‘ectogestation’). Speculation about the possibility of gestation outside the body has increased since the publication of successful animal studies demonstrating proof of concept in 2017. There are now more teams working on the prospect, and promising results of animal testing continue to be published worldwide. These devices all have a similar design—a sealed system of warm amniotic fluid (in which the subject is located), with a cannula acting as an umbilical cord to deliver oxygen and nutrients and remove waste, and a pumpless oxygenator circuit. Researchers have indicated that they hope to begin testing on human subjects in the immediate future. Equity financing for clinical translation was secured in 2022 by the team working in Philadelphia. These devices are being designed specifically as a means of improving outcomes from extremely premature birth. Artificial placentas have the potential means of enhancing reproductive decision-making by ‘taking over’ gestation for people undertaking dangerous pregnancies. The technology, in its current iteration, can facilitate only partial ectogestation (since the technology is reliant on the subject having foetal physiology). In the future, artificial placentas might have further choice-enhancing potential by offering the ability to make decisions about what degree of bodily gestational labour individuals are willing to undergo in becoming a parent in a broader range of circumstances, for example, by enabling people to opt out of performing a complete gestation, or, even of undertaking gestation at all. 

 

There is a growing body of normative ethical literature that explores the benefits and potential limitations of both UTx and artificial placentas; however, this literature can be considered somewhat disjointed. These technologies are most often explored in isolation and while it is important that we consider what is different about each technology, considering forms of assisted gestation collectively can help us see the implications of arguments made about novel technologies in their broader context. Moreover, considering the genus of assisted gestation (and thus several technologies encompassed within it) can improve our conceptual understanding of the technologies. While these technologies work in different ways, there are some obvious synergies in their potential uses to aid people who cannot gestate, and they might be thought of as alternative options to address an inability to gestate by potential service-users. Thus, in this study, multiple forms of assisted gestation were considered. The study sought to ascertain the views of reproductive rights advocates about novel forms of assisted gestation—UTx and artificial placentas—drawing together experiences and reflections connected to each. 

 

To my knowledge, before this study, there had been no empirical research conducted exploring reproductive rights advocates’ perspectives on novel forms of assisted gestation. Empirical research with individuals working with or for reproductive rights organisations is key because they have an awareness of the experiences of different people who cannot gestate or who have experienced difficulties during gestation and birth and of the political landscape. Their insights are consequently invaluable in highlighting how individuals may want to access and use novel technologies, as well as some of the challenges they may face. This article reports results from a focus group study exploring these perspectives that was analysed using a reflexive thematic analysis approach. These results have utility in helping enhance the understanding of why assisted gestation is important, and what the potential access barriers are for individuals, in further normative research.

'How Neurotech Start-Ups Envision Ethical Futures: Demarcation, Deferral, Delegation' by Sophia Knopf, Nina Frahm and Sebastian M Pfotenhauer in (2023) 29 Science and Engineering Ethics comments 

 Like many ethics debates surrounding emerging technologies, neuroethics is increasingly concerned with the private sector. Here, entrepreneurial visions and claims of how neurotechnology innovation will revolutionize society—from brain-computer-interfaces to neural enhancement and cognitive phenotyping—are confronted with public and policy concerns about the risks and ethical challenges related to such innovations. But while neuroethics frameworks have a longer track record in public sector research such as the U.S. BRAIN Initiative, much less is known about how businesses—and especially start-ups—address ethics in tech development. In this paper, we investigate how actors in the field frame and enact ethics as part of their innovative R&D processes and business models. Drawing on an empirical case study on direct-to-consumer (DTC) neurotechnology start-ups, we find that actors engage in careful boundary-work to anticipate and address public critique of their technologies, which allows them to delineate a manageable scope of their ethics integration. In particular, boundaries are drawn around four areas: the technology’s actual capability, purpose, safety and evidence-base. By drawing such lines of demarcation, we suggest that start-ups make their visions of ethical neurotechnology in society more acceptable, plausible and desirable, favoring their innovations while at the same time assigning discrete responsibilities for ethics. These visions establish a link from the present into the future, mobilizing the latter as promissory place where a technology’s benefits will materialize and to which certain ethical issues can be deferred. In turn, the present is constructed as a moment in which ethical engagement could be delegated to permissive regulatory standards and scientific authority. Our empirical tracing of the construction of ‘ethical realities’ in and by start-ups offers new inroads for ethics research and governance in tech industries beyond neurotechnology.

Neurotech

The UK Information Commissioner's Office report on neurotech states 

Neurotechnologies have continued to proliferate in the health and research sector over the past decade and may soon become part of our daily life. Our workplaces, home entertainment and wellbeing services may use neurotechnology to provide more personalised services in the years to come. 

As the UK’s data protection regulator, the Information Commissioner’s Office (ICO) aims to increase public trust in how organisations process personal information through responsible practice. We want to empower people to safely share their information and use innovative products and services that will drive our economy and our society. In our ICO25 strategy, we committed to set out our views on emerging technologies to reduce burdens on businesses, support innovation and prevent harms. 

This report specifically considers gathering, analysing and using information that is directly produced by the brain and nervous system, referred to as neurodata. This ranges from monitoring concentration levels at work, to more distant concepts such as smart prosthetics that can mimic brain patterns for greater responsivity. This report is a short introductory guide for those who wish to know more about neurotechnologies from a regulatory perspective. It does not consider the implications of neurodata inferred from broader biometric information, such as eye movements, gait or heartrate tracking. This formed part of our earlier work around biometric technologies. 

We examine the impact of neurotechnologies and neurodata and analyse their impact on privacy. We explore plausible scenarios and use cases for emerging neurotechnologies, and through these, raise the following issues:

• a significant risk of discrimination emerging in non-medical sectors such as the workplace, as complex systems and potentially inaccurate information become embedded in neurotechnology products and services. There may also be an increasing risk that unfair decisions could be made even when accurate information is used, discriminating in ways that have not previously been defined; 

• the need for people to clearly understand the technology and terminology. This enables organisations to meet their requirements for transparency, and enables people to understand their individual rights. Without this, people will be unable to provide clear consent for processing when appropriate and organisations may struggle to address the challenges of automated processing of neurodata; and 

• a need for regulatory co-operation and clarity in an area that is scientifically, ethically and legally complex.

We will address these areas of concern through:

• ongoing engagement with key stakeholders across industry, regulation, academia and civil society. This will include inviting organisations to work with our Regulatory Sandbox to engineer data protection into these technologies;  

• engagement with the public to better understand their knowledge and concerns about neurotechnologies and privacy; and 

• producing neurotechnology specific guidance in the longer term. This will address the need for regulatory clarity and set clear expectations about the responsible and compliant use of neurodata.

 

We will address some other issues elsewhere, as we build on our Artificial Intelligence (AI) Framework and forthcoming guidance on workplace surveillance. This will include potential neurodiscrimination arising through inaccurate information or inappropriate processing and decision-making.

Emotions

'Physiognomic Artificial Intelligence' by Luke Stark and Jevan Hutson  in (2022) 32  Fordham Intellectual Property, Media & Entertainment Law Journal 922   comments

The reanimation of the pseudosciences of physiognomy and phrenology at scale through computer vision and machine learning is a matter of urgent concern. This Article, which contributes to critical data studies, consumer protection law, biometric privacy law, and anti-discrimination law, endeavors to conceptualize and problematize physiognomic artificial intelligence (AI) and offer policy recommendations for state and federal lawmakers to forestall its proliferation. 

Physiognomic AI, we contend, is the practice of using computer software and related systems to infer or create hierarchies of an individual’s body composition, protected class status, perceived character, capabilities, and future social outcomes based on their physical or behavioral characteristics. Physiognomic and phrenological logics are intrinsic to the technical mechanism of computer vision applied to humans. In this Article, we observe how computer vision is a central vector for physiognomic AI technologies, unpacking how computer vision reanimates physiognomy in conception, form, and practice and the dangers this trend presents for civil liberties. 

This Article thus argues for legislative action to forestall and roll back the proliferation of physiognomic AI. To that end, we consider a potential menu of safeguards and limitations to significantly limit the deployment of physiognomic AI systems, which we hope can be used to strengthen local, state, and federal legislation. We foreground our policy discussion by proposing the abolition of physiognomic AI. From there, we posit regimes of U.S. consumer protection law, biometric privacy law, and civil rights law as vehicles for rejecting physiognomy’s digital renaissance in artificial intelligence. Specifically, we argue that physiognomic AI should be categorically rejected as oppressive and unjust. Second, we argue that lawmakers should declare physiognomic AI to be unfair and deceptive per se. Third, we argue that lawmakers should enact or expand biometric privacy laws to prohibit physiognomic AI. Fourth, we argue that lawmakers should prohibit physiognomic AI in places of public accommodation. We also observe the paucity of procedural and managerial regimes of fairness, accountability, and transparency in addressing physiognomic AI and attend to potential counterarguments in support of physiognomic AI.

Stark's 'The emotive politics of digital mood tracking' in (2020) 22(11) New Media and Society 2039-2057 comments 

A decade ago, deploying digital tools to track human emotion and mood was something of a novelty. In 2013, the Pew Research Center’s Internet & American Life Project released a report on the subject of “Tracking for Health,” exploring the growing contingent of Americans keeping count of themselves and their activities through technologies ranging from paper and pencil to digital smart phone apps (Fox and Duggan, 2013). These systems generate what Natasha Dow Schüll terms more broadly “data for life” (Schüll, 2016), traces of our everyday doings as recorded in bits and bytes. Mood tracking, about which the survey queried, received so few affirmative responses that it did not rate at even 1% of positive answers. 

Yet in the interim, emotion in the world of computational media has become big business (McStay, 2016, 2018; Stark, 2016, 2018b; Stark and Crawford, 2015). Using artificial intelligence (AI) techniques, social networks such as Twitter and Facebook have joined dedicated health-tracking applications in pioneering methods for the analysis of emotive and affective “data for life.” These mood-monitoring and affect-tracking technologies involve both active self-reporting by users (Korosec, 2014; Sundström et al., 2007) and the automated collection of behavioral data (Isomursu et al., 2007)—methods often collectively known as digital phenotyping (Jain et al., 2015), or the practice of measuring human behavior via smart phone sensors, keyboard interactions, and various other features of voice and speech (Insel, 2017). This continuum of technologies allows an analyst to extrapolate a range of information about the physiology, activity, behaviors, habits, and social interactions from everyday digital emanations (Kerr and McGill, 2007). 

The past few years have also seen policymakers and the public becoming increasingly attuned to the political impacts of digital media technologies, including AI and machine learning (ML) systems (Barocas and Selbst, 2016; Crawford and Schultz, 2013; Diakopoulos, 2016). Citizens, activists, and elected politicians are eager to address the ways in which technical particularities of such systems influence social and political outcomes via design and deployment (Buolamwini and Gebru, 2018; Dourish, 2016; Johnson, 2018). Yet critical analyses and responses to these tools of what Zuboff (2019) terms “surveillance capitalism” must account for the role of human affect, emotion, and mood in surveillance capitalism’s extraction and contestation. As Raymond Williams observed, working toward an understanding of the barriers to economic and social justice means being first and foremost “concerned with meanings and values as they are actively lived and felt” (Williams, 1977: 132). 

Here, I perform a close reading and comparative values in design (VID) analysis (Flanagan and Nissenbaum, 2014; Friedman et al., 2006) of MoodPanda and Moodscope, two popular consumer applications for tracking mood. Human emotions themselves arise from a tangled nexus of biological, cultural, and contextual factors (Boehner et al., 2005; Sengers et al., 2008). As such, I argue that the design choices in each service shape the particular dynamics of political economy, sociality, and self-fashioning available to their users, and that these design decisions are exemplary of the ties between the computational politics of surveillance capitalism (Tufekci, 2014; Zuboff, 2019), and the quantification and classification of human emotion via digital mechanisms (Stark, 2018a). 

Drawing on Tufekci (2014, 2017), Papacharissi (2014), and others (Ahmed, 2004; Martin, 2007), I articulate how the affordances of mood-tracking services such as Moodscope and MoodPanda are indexical to a broader emerging emotive politics mediated and amplified by digital systems. The dynamics of emotive politics underpin many contemporary digitally mediated sociotechnical controversies, ranging from media manipulation by extremist actors, negative polarization, and “fake news,” to collective action problems around pressing global crises such as climate change. Human passions have always been understood as an element of political life, but the particular technical and social affordances of digital systems configure these responses in particular ways: emotive politics foreground contestations regarding how we as social actors should interact together in what Papacharissi (2014) terms as “affective publics,” and the weights and ways in which we as designers, participants, and citizens should treat human feeling as dispositive features of civic discourse. Mood tracking’s explicit engagement with human emotion as a mediated, embodied state points toward how emotive politics emerge out of designer expertise, technical features, and the social contexts and practices of everyday digital mediation (Dourish, 2004; Kuutti and Bannon, 2014). 

In this analysis, I also seek to highlight the ways in which user interface and experience (UI/UX) design shape political outcomes alongside the structures of algorithms and databases (Dourish, 2016; Montfort and Bogost, 2009: 145)—though the groundbreaking work of scholars such as Johanna Drucker (2014) and Lisa Nakamura (2009) means this insight should be of surprise to no one. “The same quantitative modulations and numerical valuations required by the new information worker,” Alexander R. Galloway likewise observes, come “in a dazzling array of new cultural phenomena…to live today is to know how to use menus” (Galloway, 2006). Analyses taking interface design into account as an aspect of broader conversations around the fairness, ethics, and accountability of digital systems, which I seek to model here, will bolster the interdisciplinary work of interrogating the impact of these automated systems on our collective political future.

'The “Criminality from Face” Illusion' by Kevin W. Bowyer, Michael C. King, Walter Scheirer, and Kushal Vangara comments 

 Criminal or not? Is it possible to create an algorithm that analyzes an image of a person’s face and accurately labels the person as Criminal or Non-Criminal? Recent research tackling this problem has reported accuracy as high as 97% [14] using convolutional neural networks (CNNs). In this paper, we explain why the concept of an algorithm to compute “criminality from face,” and the high accuracies reported in recent publications, are an illusion. 

Facial analytics seek to infer something about an individual other than their identity. Facial analytics can predict, with some reasonable accuracy, things such as age [10], gender [6], race [9], facial expression / emotion [25], body mass index [5], and certain types of health conditions [29]. A few recent papers have attempted to extend facial analytics to infer criminality from face, where the task is to take a face image as input, and predict the status of the person as Criminal / Non- Criminal for output. This concept is illustrated in Figure 1. 

One of these papers states that “As expected, the state-of- the-art CNN classifier performs the best, achieving 89.51% accuracy...These highly consistent results are evidences for the validity of automated face-induced inference on criminality, despite the historical controversy surrounding the topic” [40]. Another paper states that, “the test accuracy of 97%, achieved by CNN, exceeds our expectations and is a clear indicator of the possibility to differentiate between criminals and non- criminals using their facial images” [14]. (During the review period of this paper, we were informed by one of the authors of [14] that they had agreed with the journal to retract their paper.) A press release about another paper titled “A Deep Neural Network Model to Predict Criminality Using Image Processing” stated that “With 80 percent accuracy and with no racial bias, the software can predict if someone is a criminal based solely on a picture of their face. The software is intended to help law enforcement prevent crime.” The original press release generated so much controversy that it “was removed from the website at the request of the faculty involved” and replaced by a statement meant to defuse the situation: “The faculty are updating the paper to address concerns raised” [13]. 

Section II of this paper explains why the concept of an algorithm to compute criminality from face is an illusion. A useful solution to any general version of the problem is impossible. Sections III and IV explain how the impressive reported accuracy levels are readily accounted for by inadequate experimental design that has extraneous factors confounded with the Criminal / Non-Criminal labeling of images. Learning incidental properties of datasets rather than the intended concept is a well-known problem in computer vision. Section V explains how Psychology research on first impressions of a face image has been mis-interpreted as suggesting that it is possible to accurately characterize true qualities of a person. Section VI briefly discusses the legacy of the Positivist School of criminology. Lastly, Section VII describes why the belief in the illusion of a criminality-from- face algorithm potentially has large, negative consequences for society.

Neurosurveillance

'Neuro-Surveillance and the Right to be Human at Work' by Valerio De Stefano in On Labor comments

In 2018, the news reported that some assembly line workers had been asked to wear caps that monitor brain waves in order for managers to adjust the pace of production and workflows. Some observers raised doubts on the reliability of these tools, and even cast doubts on their actual functioning, but it is undeniable that forms of mental surveillance are increasingly coming to our workplaces. “Sociometric badges”, wearable tools tracking emotions and stress by collecting data on heartbeats and the tone of voice, for instance, are spreading in the United States. 

In 2018, the news reported that some assembly line workers had been asked to wear caps that monitor brain waves in order for managers to adjust the pace of production and workflows. Some observers raised doubts on the reliability of these tools, and even cast doubts on their actual functioning, […] 

Most of these practices should be urgently restricted. Losing one’s mental privacy arguably threatens one of the core elements of being human. If this occurs at the workplace, where workers are already subject to quasi-dictatorial managerial prerogatives, the consequences could be disastrous. Yet, no significant attention has been devoted to how neurotechnologies, and other forms of mental surveillance, may impact on workplaces. 

Besides brain waves monitoring and emotional tracking, facial scans are also widely used in recruitment, with artificial intelligence analyzing “how a person’s face moves” to detect ”how excited someone seems about a certain work task or how they would behave around angry customers”. Research projects exploring how to connect brains to technological devices are underway that could have detrimental consequences for workers. 

The more these practices progress, the more labour scholars should raise concerns about them. Experimentations and implementation of these practices need data, and workplaces are perfect data mines. If regulation is not brought up-to-speed, the future world of work risks being one where employers require employees to use tools that collect data on their brain activity to engage in a unrestrained quest for productivity, to predict their behavior and even to monetize on their data by making them accessible to third parties. 

Given the imbalance between employers and workers, there is limited possibility for workers to refuse such surveillance without risking to lose their jobs. This is why European countries have promoted governance of these practices through collective bargaining and workers’ representatives’ involvement. The EU General Data Protection Regulation provides that EU Member States may introduce, by law or by collective agreements, “specific rules to ensure the protection of the rights and freedoms in respect of the processing of employees’ personal data in the employment context”. 

Much more heed, however, needs to be paid to “neuro-surveillance at work”. Neuroscientist Marcello Ienca, for instance, called for the recognition of new human rights to face the rise of neurotechnology, including rights to mental privacy and integrity.

Neurolaw and Privacy

'Will There Be a Neurolaw Revolution?' by Adam J. Kolber in (2014) 89 Indiana Law Journal 807-845 argues that

The central debate in the field of neurolaw has focused on two claims. Joshua Greene and Jonathan Cohen argue that we do not have free will and that advances in neuroscience will eventually lead us to stop blaming people for their actions. Stephen Morse, by contrast, argues that we have free will and that the kind of advances Greene and Cohen envision will not and should not affect the law. I argue that neither side has persuasively made the case for or against a revolution in the way the law treats responsibility. 

There will, however, be a neurolaw revolution of a different sort. It will not necessarily arise from radical changes in our beliefs about criminal responsibility but from a wave of new brain technologies that will change society and the law in many ways, three of which I describe here: First, as new methods of brain imaging improve our ability to measure distress, the law will ease limitations on recoveries for emotional injuries. Second, as neuroimaging gives us better methods of inferring people’s thoughts, we will have more laws to protect thought privacy but less actual thought privacy. Finally, improvements in artificial intelligence will systematically change how law is written and interpreted. … 

The emerging field of neurolaw addresses two major topics that have only limited overlap. The “neurolaw of responsibility” concerns how neuroscience will and should affect laws related to responsible action. It was traditionally addressed by punishment theory and the philosophy of action. The “neurolaw of technology,” by contrast, concerns the ways the law will and should respond to new brain-related technologies. It covers issues traditionally addressed by applied ethics. Both topics require familiarity with law and neuroscience, but they otherwise examine rather different issues. Nevertheless, since both fields happen to involve law and neuroscience, the neurolaw moniker seems to have stuck. 

Greene, Cohen, and Morse write principally about the neurolaw of responsibility. They spend much of their energy defending their substantive views about free will, though none of them purport to offer a new argument to break the free will impasse. Greene and Cohen also claim that advances in neuroscience will change the way we think about punishment, but they have yet to persuasively defend the claim. Similarly, Morse may be right that we ought to understand the law in compatibilist terms, but current law may be rooted in contrary assumptions. 

While prospects for a responsibility revolution remain hard to predict, I claim that there will be a technology-driven neurolaw revolution. The law will change in many ways, and I focus on three hypotheses: (1) the differences in how the law treats emotional and physical injuries will diminish as neuroscientists develop more objective methods of identifying and assessing emotional injuries; (2) new methods of “mind reading” will lead us to have less thought privacy but more thought privacy laws; and (3) as autonomous and semiautonomous machines become more integrated into human life, they will have systematic effects on the law and its interpretation, perhaps by increasing the concretization of the law. The precise details of how technology will develop are hard to predict, but by trying to predict the path of technology, we can hope to make the law better prepared for the changes to come.

Kolber argues that there will be "More Privacy Laws but Less Privacy" -

 Researchers are working on a variety of technologies aimed at what can loosely be referred to as mind reading. For example, based on measurements of brain activity, researchers can make pretty good guesses about what images are shown to a subject in a brain scanner, be it a still image or even, to some extent, a video. One recent study demonstrated that subjects under fMRI can be taught to mentally spell words in a manner that can be decoded in real time by researchers, a technique that could prove especially helpful for people with locked-in syndrome or other conditions that make it difficult to communicate. Neuroscientist Jack Gallant predicts that “[w]ithin a few years, we will be able to determine someone’s natural language thoughts using fMRI-based technology.” 

These new brain imaging techniques point to a future where our thoughts will not be as private as they are now. We will not read minds directly in any spooky sense, but we will continue to get better at identifying correlations between brain activity and mental activity and using brain activity to make predictions about mental activity. 

Legal scholars have focused their attention on efforts to develop more accurate lie detectors. Brain-based methods of deception detection are still in the early stages. Much of the research compares the brain activity of a group of “honest” subjects relative to a group of “dishonest” subjects. More helpful research to determine whether a particular person is lying is beginning to accumulate, but the testing has always been done in somewhat artificial contexts. If we put aside concerns about how well these experiments apply to real-life contexts, most published studies report using fMRI to distinguish honesty and deception at accuracies “between 70% and slightly over 90%.” 

But even if we develop a lie detector that works well with the cooperative subjects that tend to participate in experiments, very little research examines the possible countermeasures people could take to fool such a device. One fMRI study was 100% accurate in detecting the lies of individual subjects, but accuracy fell to 33% when subjects used countermeasures they were trained to apply. So even though at least two companies have marketed brain-based methods of lie detection, many neuroscientists are skeptical of the current state of the technology. Indeed, two recent attempts to introduce fMRI evidence of deception in court were unsuccessful. 

Nevertheless, deception detection has so many potential uses that the incentives to improve it are quite strong. Someday, the technology will at least be a useful aid in assessing credibility. When that day comes, many questions will be raised about how, if at all, the technology should be used in court. The real question we ought to ask ourselves when considering some supposed lie detector is: will we tend to get more accurate outcomes with or without it? 

The answer may depend on the context. Lie detection evidence offered by prosecutors to provide evidence of guilt beyond a reasonable doubt would have to be extremely accurate, while lie detection evidence offered by a defendant to generate a reasonable doubt could be much more imperfect.  Deciding whether or not brain-based lie detection will improve outcomes, however, will put us in an awkward position: we will have to compare the error rates of lie detection technology to our current technology, namely, the jury, and we know relatively little about how well juries assess credibility. What we do know is that people are not very good at detecting deception, and there is little correlation between people’s confidence in their ability to detect deception and their accuracy. Our entrenched preference for jury decision making is largely a result of the path of history, rather than an empirically validated conclusion about how good juries are at discerning credibility. 

In an opinion in United States v. Scheffer, Justice Clarence Thomas, joined by three other Justices, wrote that a rule banning all polygraph evidence in military trials serves the legitimate government interest of preserving jurors’ “core function of making credibility determinations in criminal trials.” According to Thomas, “[a] fundamental premise of our criminal trial system is that ‘the jury is the lie detector.’” His remarks admit the possibility that even perfectly accurate lie-detection evidence could be excluded from the courtroom on the ground that it would infringe the province of the jury. 

In my view, excluding accurate lie-detection information to protect the province of the jury makes a mockery of the justice system. The most important role of trials is to uncover the truth as best we can. To do so, we ought to use the best technology that cost-effectively helps us do so. There are legitimate concerns that poor quality lie detection evidence could irrationally sway jurors. They may not understand how the technology works or how to interpret known rates of error. But it would be foolish to keep some high-quality future lie detector out of the courtroom — under a blanket rule — simply because credibility determinations have traditionally been made by jurors. 

Of course, even a perfectly accurate lie detector could not usurp all jury functions. Some cases do not depend on credibility assessments at all. For example, whether or not conduct was consistent with that of a reasonably prudent person cannot be determined by a lie detector alone. Moreover, when cases do depend on witness credibility, there is an important difference between honesty and truth. Honest assertions are not necessarily true. A person may believe he committed a crime that, in fact, never occurred. Similarly, a dishonest assertion can turn out to be true. A gunman may believe he fired the coup de grâce shot that ended the life of a rival gang member. Denying that he killed the rival would be dishonest, even if unbeknownst to him, the deceased was already dead before he fired. 

If direct attempts at brain-based lie detection fail, other mind reading efforts may still prove helpful. The technologies discussed in the preceding section on the experiential future can serve as indirect methods of lie detection by telling us whether a person’s pain claims are likely to be false. (In fact, pain measurement techniques could give us information that cannot be obtained from truthful subjects. Even when a person honestly reports his pain as “9” on a scale of 1 to 10, we cannot easily compare his report to those of others.) 

Researchers are improving their understanding of other experiences, as well, including sexual arousal. One study examined the brain activity of male pedophiles and male non-pedophiles when shown images of nude children. Researchers used brain activity to accurately classify the pedophilia status of more than 90% of subjects. While this technique may be subject to countermeasures, it may be less so than other techniques used to classify pedophiles. Another study looked at the brain activity of subjects while they looked at male and female human genitalia. Researchers could determine sexual orientation with more than 85 percent accuracy. 

Of course, all of this work on mind reading raises privacy concerns. The pedophilia research suggests that fMRI could someday be used to assess the likelihood that a person has committed or will commit a sex crime. The research on sexual orientation could potentially bear on the distribution of assets in a divorce or the way prisoners are segregated. Other neuroscientific research may uncover conscious or unconscious racial biases. People could be scanned for one purpose, say, to see how an advertising campaign affects their brains, while they inadvertently generate information that bears on their racial biases, sexual orientation, and other sexual preferences. One group of researchers recently demonstrated that the very simple electroencephalography (EEG) sensors in certain mass-market video games can already be used to make plausible inferences about gamers’ private “information related to credit cards, PIN numbers, the persons known to the user, or the user’s area of residence,” and may enable more confident inferences as these sensors improve. 

Brain imaging may even inform questions about mens rea. It might help us assess a person’s capacity to generate some mental state or bear on the credibility of a person’s statements about his past mental states. Brain imaging might even have more direct applications. For example, one group of researchers is trying to use fMRI to identify the culpable mental states described by the Model Penal Code. Imagine a border crossing where someone is transporting a suspicious container. Before opening it, we could scan the brain of the person carrying the container to see if his brain is consistent with a culpable mental state of knowledge, recklessness, or negligence with respect to its contents. (The person might have to believe he was randomly selected for screening so that the mere fact of being selected does not significantly alter his beliefs.) 

Accurate mind-reading technologies would raise a host of questions: For example, when, if ever, could prosecutors use brain-based lie detectors to incriminate or defendants to exculpate? How would police and other investigators use such tools? Could they be used by employers to make hiring and firing decisions? 

Even if accurate mind-reading techniques are still decades away, we already have reason to think about their implications because of what I call the technological look-back principle. If we develop an accurate lie detector thirty years from now, you can be asked in 2044 about your conduct today in 2014. When you are in such a scanner in 2044, your spouse could ask if you have ever been unfaithful, and the police could ask if you have ever killed someone. And just as campaigning politicians often make their tax returns public even though they are under no legal obligation to do so, voters may expect politicians to go into a scanner and tell them what their intentions really are and whether or not they have ever acted corruptly. 

I am not arguing that we need legislation today to prepare for all of the potential future uses of mind-reading techniques.  We would have little confidence that such legislation would survive the intervening period or that it would it take the appropriate form. Moreover, we often worry too much about the privacy concerns raised by new technologies in ways that unnecessarily hinder their development. 

But those expecting to be alive in coming decades or who care about those who will should begin to think about the privacy implications of mind-reading technologies. Many who shed their DNA while committing crimes before DNA sequencing became common are now in prison, prosecuted with evidence they never imagined could be used against them. Our memories may become the evidence that embarrasses or incriminates us in the future. 

I offer two general predictions about how our rights to privacy may change in a world with better mind-reading technology. First, as the preceding suggests, we will have less mental privacy as advances in neuroscience make it easier to infer thoughts and thought patterns. We strike a balance between the societal value of making information public and the value to a person or group of people of keeping it private. These costs and benefits push and pull each other to reach a certain equilibrium. Neuroscience will reduce the costs of obtaining otherwise private information and will likely enable access to information that would otherwise be unavailable. Given that societal demand for information is likely to stay the same or increase, the equilibrium is likely to shift toward more information gathering. 

In the days before the Internet, one could hire a private investigator to learn about people’s occupations, family members, and various likes and dislikes. Today, such information is frequently easy to obtain. Indeed, many people publicize it themselves on social networking sites. Even when people try to keep their own information private, their associates still generate information about them. As technology makes information easier to obtain, it becomes harder to keep private. Second, I speculate that we will craft more laws to protect thought privacy. Right now, there is little we can do to penetrate the thoughts of people who prefer to keep them secret. Only when we have plausible methods of doing so will we fully see the need to create laws to protect thought privacy. For example, as polygraphs became more reliable and widespread, Congress passed the Federal Employee Polygraph Protection Act in 1988 to prohibit most private employers from subjecting employees to polygraphs and other forms of lie detection. And just as we have seen an onslaught of laws to protect electronic privacy, we will see new laws directed at protecting the privacy of our thoughts. Laws addressing rights to read minds or to be free of mind reading will grow more prevalent, complex, and controversial in a world with more accurate neurotechnology. Hence, we will have more law protecting thought privacy but less actual thought privacy.

CyberAttacks and Neurotech Ethics

'Ghost in the Network' by Derek Bambauer - forthcoming in (2014) 162 University of Pennsylvania Law Review indicates that

Cyberattacks are inevitable and widespread. Existing scholarship on cyberespionage and cyberwar is undermined by its futile obsession with preventing attacks. This Article draws on research in normal accident theory and complex system design to argue that successful attacks are unavoidable. Cybersecurity must focus on mitigating breaches rather than preventing them.

First, the Article analyzes cybersecurity’s market failures and information asymmetries. It argues that these economic and structural factors necessitate greater regulation, particularly given the abject failures of alternative approaches. Second, the Article divides cyber-threats into two categories: known and unknown. To reduce the impact of known threats with identified fixes, the federal government should combine funding and legal mandates to push firms to redesign their computer systems. Redesign should follow two principles: disaggregation, dispersing data across many locations; and heterogeneity, running those disaggregated components on variegated software and hardware. For unknown threats -- “zero-day” attacks -- regulation should seek to increase the government’s access to markets for these exploits. Regulation cannot exorcise the ghost in the network, but it can contain the damage it causes.

Bambauer argues that

While a complete defense to zero-day attacks is impossible, policymakers can improve cybersecurity with three regulatory moves: mandatory access to public zero-day markets for the federal government, required confidential reporting on transactions by firms in those markets, and a reward system for researchers who share vulnerabilities with the government. Congress should pass legislation implementing these measures. America should try to convert unknown unknowns to known unknowns. First, firms that transact in software security vulnerabilities should be required to permit the federal government to participate in any offerings or services they provide, on non-discriminatory terms. If Vupen, for example, sought to sell zero-day exploits to France’s security services, but not to America’s National Security Agency, that would be problematic. American law should make paid access by the U.S. government a condition of legal operation for software security firms. This enables the government to develop and deploy countermeasures to at least some zero-day attacks.

Congress has taken analogous measures for other potential risks to national security. For example, one cannot obtain a patent for inventions in nuclear materials or weapons. Such inventions are eligible for a governmental reward scheme, but not for patents. And, the statute transfers rights to the invention from the inventor to the federal government. Similarly, export controls restrict private firms’ ability to engage in transactions with foreign countries. One may not transfer software utilizing encryption to countries such as Iran or North Korea, and one may not sell certain supercomputers to countries such as China or Russia. These rules apply to all firms within U.S. jurisdiction. Thus, Congress has either mandated or forbidden certain transactions based on national security concerns, and could mount a similar effort for zero-day sales.

Not all zero-day merchants fall under American jurisdiction, or enforcement. However, even those operating abroad likely have contacts with the United States. Vupen’s employees visit the United States. Many, if not all, such firms use financial or payment processing companies that are subject to U.S. regulation. These links provide potential leverage. Congress could attach provisions to this legislation that would allow the executive branch to designate firms that do not provide access to the government, and to require banks and payment processors to forgo transactions with them. Analogous measures have been implemented to interdict financing for terrorist groups489, and have been proposed to deal with sites offering prescription drugs or copyrighted works illegally.

Second, Congress should mandate a transaction-reporting system for firms trading in vulnerabilities. These companies should have to report, on a confidential basis, the purchaser’s identity in all transactions of zero-day exploits to the National Security Agency (NSA). This data would remain confidential, and should be designated as statutorily immune from discovery or other use unless the NSA expressly chooses to share it. The statute should enable auditing of firms’ records by the NSA if the agency is able to demonstrate an objectively reasonable basis to suspect inaccuracies or falsification. To make this provision less objectionable for the vulnerability merchants, Congress should include payments to firms that report. While additional spending is politically difficult, this expenditure would be a small but worthwhile investment in security.

Similar reporting systems are widely used to mitigate risk. The National Aeronautics and Space Administration encourages confidential reporting of “near miss” incidents – those that nearly resulted in aviation mishaps – to improve safety procedures and detect product defects. Insurers offering policies for medical malpractice liability must report judgments and settlements to the National Health Practitioner Data Bank. This malpractice information is available for use by state medical licensing boards and federal agencies, but is otherwise confidential. The Federal Railroad Administration is testing a Close Calls Demonstration Project to identify risks in rail operations via confidential reporting of near-miss incidents. The Department of Veterans Affairs has a similar system for patient safety, as does the Federal Communications Commission for network outages.

A zero-day reporting system has several benefits. It would enable the government to detect problematic sales, particularly to unfriendly states and to insecure parties. It would increase the effectiveness of countermeasures that mitigate zero-day exploits by providing a rough guide to how widely distributed a particular attack tool is. It would allow the government to identify whether firms follow their stated criteria for sales (such as Vupen’s self-imposed limit to NATO countries and clients), and to scrutinize suspect firms more closely. Lastly, it would provide a crude estimate of the ebb and flow of the zero-day threat, and to the platforms and applications viewed by the merchant as worthy of attention (and payment).

Finally, Congress should authorize a “bug bounty” program. Its goal would be to collect zero-day exploits, and to encourage researchers to sell their findings to the U.S. government rather than to private firms or other nation-states. A government agency, such as the NSA or the U.S. Computer Emergency Readiness Team, should be provided funds to buy zero-day vulnerability information. The entity selling the exploit, such as a security research firm, would have to certify under penalty of perjury that it had not previously shared the vulnerability information with others, and would have to agree contractually not to do so in the future. Congress should consider backing these requirements with substantial criminal penalties. Arms dealers who sell to both sides are held in low esteem.

Similar private bounty programs, such as by Google and Mozilla, have had considerable success in identifying and remediating bugs. The funding, and amount paid per bug, should be generous: removing zero-days from the Internet ecosystem is highly worthwhile. Moreover, generous payments will have two further beneficial effects. First, it will spur researchers to search for additional bugs. These bugs are like latent defects in a product – they lurk, creating risk, until discovered. Second, paying above-market rates makes it more difficult for others to purchase zero-days. Pushing others out of the zero-day market is useful both offensively and defensively. Offensively, accumulating zero-days provides the U.S. with the building blocks for future Stuxnets. Defensively, it reduces the likelihood that American firms or government entities will fall vulnerable to attacks.

The bug bounty program will create several challenges. First, price: more competition for zero-day exploits will drive up their cost. This increase will burden the public fisc slightly, but helpfully generates added incentives for research into bugs. Second, the government will need to decide how to use exploit information. Congress could establish rules for what NSA may do with the data, or it could defer to the agency (and, by extension, the executive branch) to make that decision. If the NSA uses the exploits to build cyberweapons, such as Stuxnet, or to enable others to do so, it is likely to share vulnerability information less widely than it would without a vision of offensive use. If the agency enables other government entities or private firms to take precautions against the zero-days, it risks having those patches shared, including with potential targets. And, there is an ironic feedback effect: the more important the vulnerability, the greater the temptation to weaponize it, and thus to withhold it from other affected parties.

The hardest decision regarding sharing is determining whether to notify the affected vendor. This Article argues that telling the vendor about the vulnerable code should be the default practice, with two caveats. First, the NSA should work with the vendor to ensure the patch for the vulnerability is maximally effective and minimally visible. If the company draws attention to the patch’s criticality, it may signal to anyone who has independently discovered it that the window of vulnerability is closing – which could draw attacks. Second, NSA should work with the vendor to include detection code in patches. This would help the agency estimate how often vulnerabilities are discovered independently, and perhaps to detect double-dealing by researchers participating in the bug bounty system.

This Article’s solutions for the zero-day problem – the unknown unknowns – differ in character from those for vulnerabilities with existing solutions (the known unknowns) in that they have a greater focus on prevention. Mitigation is still invaluable: disaggregation and heterogeneity are just as helpful for zero-days as for known bugs. However, preventive steps are more important for zero-day exploits. With known vulnerabilities, defenses are possible, though logistically constrained by externalities, information costs, and system complexity. With zero-days, defenses are impossible. Defenders must rely solely on mitigation and recovery. And while prevention tends to be overrated in cybersecurity literature, it remains useful. In particular, even if complete prevention is impossible, defenders may be able to reduce an exploit’s effects – for example, by allowing a server to terminate an affected program, rather than having it cause the server to crash. This is similar to a public health approach: even if one cannot prevent people from contracting a virus, we may be able to make it less lethal. Thus, the three-part agenda above seeks to increase America’s access to information about zero-days, thereby enabling precautions and improving mitigation.

'The Representations of Novel Neurotechnologies in Social Media: Five Case Studies' by Allyson Purcell-Davis in (2013) 19(1) The New Bioethics 30 comments 

The aim of this study was to conduct an analysis of how certain novel neurotechnologies are represented and communicated within social media. The research was conducted as part of a report called Novel Neurotechnologies: Intervening in the Brain and was initially commissioned by the Nuffield Council on Bioethics. Before producing the final report, a working party examined the ethical, social and legal issues surrounding the use of novel neurotechnologies in both therapeutic and non-clinical settings, with the objective of providing an ethical framework to guide those practices. This ethical framework includes the desire to see responsible communication of novel neurotechnologies within the media. 

Chapter 9 of the report, ‘Communication of Research and the Media,’ examines issues raised in the reporting of research into, and the development of, the uses of novel neurotechnologies. As part of the findings contained within this chapter, research was undertaken to provide a ‘snapshot’ of the kinds of representations found within social media: the nature of the connections between users; the kinds of messages that are uploaded; how media content is used; and the kinds of representations of user groups found within posts. This ‘snapshot’ is presented using five case studies