1. Introduction
At the Code Conference in 2016, the CEO of Tesla, Elon Musk, suggested that humanity would ‘be left behind’ by Artificial Intelligence.Footnote 1 He predicted a ‘benign scenario’, stating that ‘[w]e would be so far below them [robots] in intelligence that we would be like a pet’.Footnote 2 As a solution he recommended the ‘merger of biological intelligence and machine intelligence’, the ‘symbiosis’ of humanity with machines through a ‘neural lace’ – that is, brain-machine interfaces (BMIs), also called ‘brain-computer interfaces’ (BCIs).Footnote 3
Professor Jacques Vidal introduced the term ‘brain-computer interfaces’ in 1973.Footnote 4 He postulated that electrical brain signals could be used in man-computer communication and may allow control over external devices, such as prosthetic devices or spaceships.Footnote 5 Neuroscientists worked for years to enable people who were suffering from disabilities or disorders to communicate or move again.Footnote 6 Progress in science has indeed revealed that:Footnote 7
every action our body performs begins with a thought and with every thought comes an electrical signal, [and that this signal] can be received by the brain-computer interface, consisting of an electroencephalograph (EEG) or an implanted electrode, which can then be translated, and then sent to the performing hardware to produce the desired action.
However, it has appeared progressively that another use could be made of neuroscience: namely, the augmentation of human capacities, or human enhancement, as now advocated by Elon Musk. In particular, this application has received the attention of several national armed forces, who aspire for military superiority through the development and use of BCIs, and are currently developing or planning to develop such technologies.Footnote 8 For instance, research sponsored by the US Defence Advanced Research Projects Agency (DARPA) aims to develop ‘telepresence’,Footnote 9 thus replacing ground troops and jet fighters by mind-controlled avatarsFootnote 10 and drones,Footnote 11 and is focusing on threat-detection systems that analyse brain signals.Footnote 12
For the purpose of this article, human enhancement is defined as the use of medical techniques ‘to improve human form or functioning beyond what is necessary to sustain or restore good health’,Footnote 13 or to depart ‘from species-typical normal functional organization or functioning’.Footnote 14 If bioethicists have debated the desirability of human enhancement and expressed concerns over the risk of eugenics since the 1970s, with the emergence of genetic engineering,Footnote 15 interest from states,Footnote 16 international organisationsFootnote 17 and international lawyers on this question is recent.Footnote 18 However, their work tends to focus on the employment of human enhancement. In contrast, the purpose of this article is twofold. First, it aims to identify the rules of international law that apply to experimentation – namely, what happens in laboratories. Second, it aims to assess the lawfulness of experimentation with BCIs in light of those rules. In fact, several techniques may be described as human enhancement when they are used for the sole purpose of improving performance: these include resort to surgery (for example, Lasik),Footnote 19 medicinal substances (for example, Ritalin),Footnote 20 prosthetics and cybernetics. If some rules may apply equally to different forms of enhancement – such as the prohibition of experimentation on protected persons, which stems from international humanitarian law, and the requirement of free consent, which appears in international human rights law – cybernetical enhancement raises specific legal challenges, for instance, in terms of risk/benefit assessment. These devices indeed come with a unique set of risks for the subject, such as irreversible effects of intra-cortical trauma, haemorrhage, and infection,Footnote 21 most notably because they often involve direct and physical interferences with the brain or the skull. In fact, three main types of BCI exist. Non-invasive devices are less risky because they ‘rest outside the brain and try to capture brain signals’,Footnote 22 usually through scalp recording (EEG).Footnote 23 In contrast, with invasive devices ‘electrode array heads are buried within the brain itself on a permanent basis’.Footnote 24 With semi-invasive devices, for their part, electrodes ‘are implanted inside the skull, but rest outside the brain rather than within the gray matter’.Footnote 25 At the same time, they have huge potential for helping disabled and sick persons on a daily basis, and may reduce mortality on the battlefield.
Against this background, the article is structured as follows. In the next section I explain that the rules of international humanitarian law adopted in the aftermath of the Second World War, in reaction to the macabre experiences on inmates in Nazi concentration camps,Footnote 26 make personal therapeutic benefit an essential condition for the legality of experiments on civilians and prisoners of war (Section 2). In the third section I study international human rights law, and I explain that two principles indirectly affect the development of enhancement technologies: (i) free consent, and (ii) the adequacy of risk and benefit – also called ‘proportionality’ (Section 3). The article then addresses the unique set of circumstances that exist when members of the armed forces are subjected to such experimentations and how those rules may be adapted to them (Section 4). A conclusion ends the article (Section 5).
At this stage of the introduction it is worth mentioning that the field of biomedical research has long been characterised by the omnipresence of non-binding instruments, which had significant influence on state domestic laws. For instance, requirements with regard to the scientific value and integrity of the project,Footnote 27 the absence of an alternative method of human testing,Footnote 28 the primacy of the individual's interest over those of society,Footnote 29 as well as appropriately qualified investigatorsFootnote 30 were already mentioned in the Nuremberg Code (1947).Footnote 31 The approval of and supervision by an ethics committee, which are now required by many states for medical research projects involving human subjects,Footnote 32 appeared in the revised version of the Declaration of Helsinki (DoH).Footnote 33 The principles of ‘respect’, ‘beneficence’ and ‘justice’ were identified by the Belmont Report (1979).Footnote 34 In fact, it appears that, among all those principles mentioned in ethical and non-binding instruments, only free consentFootnote 35 and proportionalityFootnote 36 received general consecration in positive international law. Different caveats must therefore be imposed. I refer regularly throughout the article to the measures adopted by states at the domestic level. However, regulation is imposed through different normative means and these measures do not all carry equal normative power. The legal framework in continental Europe is based on legislation that is binding for all institutions that carry out experimentation on human subjects. In Canada, Australia and the US, the legal framework is based on guidelines and regulations that are binding on research projects in receipt of public funding while other institutions are often invited to follow them.Footnote 37 In South Africa, the National Health Act (NHA) contains legal rules and establishes the National Health Research Ethics Council, which must ‘set norms and standards for conducting research on humans and animals’.Footnote 38 A further limitation of the present analysis is that because of space constraint, two aspects of international law are not addressed in the article, even though they may have relevance in terms of experimentation: these are (i) the right to privacy, and (ii) arms control law.Footnote 39
2. International Humanitarian Law: The Lawfulness of Experiments on Protected Persons Conditioned upon a Personal Therapeutic Benefit
Following the abuses of the Second World War, states took steps to prevent unethical experiments. In accordance with Geneva Convention III, ‘[p]risoners of war must at all times be humanely treated’, and acts ‘causing death or seriously endangering the[ir] health’ are prohibited.Footnote 40 In particular, they cannot ‘be subjected to physical mutilation or to medical or scientific experiments of any kind which are not justified by the medical, dental or hospital treatment of the prisoner concerned and carried out in his interest’.Footnote 41 In accordance with Geneva Convention IV,Footnote 42 protected persons should not be subject to ‘any measure of such a character as to cause the physical suffering or extermination of protected persons in their hands’ – which includes ‘medical or scientific experiments not necessitated by the medical treatment of a protected person’.Footnote 43 In addition, states have pledged ‘to enact any legislation necessary to provide effective penal sanctions’ for persons who order or perpetrate ‘grave breaches’ with, among others, ‘torture or inhuman treatment, including biological experiments’.Footnote 44 The protection of ‘persons who are in the power of the adverse Party or who are interned, detained or otherwise deprived of liberty’ was reinforced by Additional Protocol I,Footnote 45 which prohibited ‘any unjustified act or omission’ directed against their ‘physical or mental health and integrity’.Footnote 46 Accordingly, it is forbidden to subject anyone ‘to any medical procedure which is not indicated by the state of health of the person concerned’.Footnote 47 Reference is expressly made to the prohibition, ‘even with their consent’, of ‘physical mutilations’, ‘medical or scientific experiments’ and the ‘removal of tissue or organs for transplantation’.Footnote 48 Article 5(2)(e) of Additional Protocol II gave similar protection to persons detained in the framework of non-international armed conflicts.Footnote 49 It must be mentioned that Geneva Conventions III and IV have received universal acceptance (with 196 state parties), while Additional Protocols I and II have also received considerable acceptance (with respectively 174 and 169 state parties).
Under this regime, protected persons consequently may not be subject to experiments that lack a personal medical benefit, even if they consent. This interpretation is confirmed by the UKFootnote 50 and US manuals.Footnote 51 Hence, wartime experiments driven by the development of enhancement techniques would, under international humanitarian law, be illegal when carried out on prisoners of war and protected persons. It would be the same for other therapeutic experiments if they are not carried out in the interests of the subject. In addition, Additional Protocols I and II mention that any medical procedure must comply with ‘generally accepted medical standards’.Footnote 52
3. International Human Rights Law: Indirect Regulation of Experimental Enhancement through the Rules of ‘Free Consent’ and ‘Proportionality’
A first rule, which conditions the legality of biomedical research, is that of free consent. This requirement naturally applies to experimental enhancement (3.1). While obtaining free consent is a sine qua non condition for the legality of any experimentation on human beings, it is not sufficient. The lawfulness of experimental enhancement will also be constrained by a further rule: that of proportionality (3.2).
3.1. The Lawfulness of Experimental Enhancement Conditioned upon Free Consent
A fundamental rule of human rights law, which has been consecrated at the global and regional levels, is that of free consent in the framework of medical or scientific experimentation. It expressly appears in Article 7 of the International Covenant on Civil and Political Rights (ICCPR) as ‘[n]o one shall be subjected to torture or to cruel, inhuman or degrading treatment or punishment. In particular, no one shall be subjected without his free consent to medical or scientific experimentation’.Footnote 53 In fact, the lack of free consent ‘is to be considered as a sign of the inhuman character of a medical experiment’.Footnote 54 With regard to ‘persons not capable of giving valid consent’, including ‘those under any form of detention or imprisonment’, the Human Rights Committee (HRC) observed that they ‘should not be subjected to any medical or scientific experimentation that may be detrimental to their health’.Footnote 55 In addition, it observed that:Footnote 56
[w]hen there is doubt as to the ability of a person or a category of persons to give such consent … the only experimental treatment compatible with article 7 would be treatment chosen as the most appropriate to meet the medical needs of the individual.
The requirement of free consent was then adopted at the European level through the case law of the European Commission and the European Court of Human Rights (ECtHR). The Commission has long acknowledged that ‘medical treatment of an experimental character and without the consent of the person involved may under certain circumstances be regarded as prohibited by Article 3 of the [European] Convention [for the Protection of Human Rights (ECHR)]’.Footnote 57 According to Article 3 of the ECHR, ‘[n]o one shall be subjected to torture or to inhuman or degrading treatment or punishment’;Footnote 58 yet the Court also found a ‘general rule’ according to which ‘a measure which is a therapeutic necessity cannot be regarded as inhuman or degrading’.Footnote 59 The Court verifies that ‘a medical necessity has been convincingly shown to exist and that procedural guarantees for the decision exist and are complied with’.Footnote 60 For instance, the Court found in VC v Slovakia that a sterilisation ‘carried out without the informed consent of a mentally competent adult’ was contrary to Article 3 of the ECHR, even if ‘there [was] no indication that the medical staff [had] acted with the intention of ill-treating the applicant’.Footnote 61 Twenty-nine states also ratified the Convention on Human Rights and Biomedicine, known as the Oviedo Convention, which makes free consent compulsory for ‘[a]n intervention in the health field’.Footnote 62 It may be noted that the Oviedo Convention is open for accession by both member states of the Council of Europe and non-member states.Footnote 63
If it is obvious that constraining healthy persons to undergo experimental enhancement would be illegal under each of these instruments, ensuring compliance with this rule among the armed forces of a state may present particular difficulties; I will return to this aspect later. Unsurprisingly, many states in their domestic laws require the informed and free consent of the subject, often in writing.Footnote 64 The prohibition of non-consensual medical experimentation was even considered by some courts in the US as part of the ‘law of nations’,Footnote 65 the ‘law of humanity’,Footnote 66 or ‘a universally accepted norm of customary international law’.Footnote 67 The laws of several states even went further by including another rule, which stems from the Nuremberg CodeFootnote 68 and was also incorporated in the Oviedo Convention:Footnote 69 the possibility for the subjects to withdraw consent at any time and, consequently, to discontinue their participation at will.Footnote 70
3.2. Proportionality as a Criterion for the Lawfulness of Experimental Enhancement
Obtaining the consent of the subject does not give carte blanche to scientists to carry out any type of experimentation. To be lawful, the experimentation is required to satisfy a further condition: proportionality (the adequacy of risk and benefit).Footnote 71 This condition is obvious in the ICCPR framework (3.2.1), but the situation is more complex in the framework of the Council of Europe (3.2.2).
3.2.1. The ICCPR Framework
As previously underlined, Article 7 of the ICCPR prohibits acts of torture, as well as cruel, inhuman or degrading treatment or punishment, and highlights that no one shall be subjected to medical or scientific experimentation ‘without his free consent’. However, two questions are of importance in the field of experimentation. First, could a subject consent to any type of risk? Second, is an experimentation still lawful when the expected benefit for the subject is great, but comes with significant risk?
With regard to the first question, it is worth underlining that during the travaux préparatoires, the retention of the words ‘without his free consent’ was criticised:Footnote 72
It was argued that the words were not only redundant, but might open the door to abuses in that it would be possible to justify experimentation of a criminal nature on the pretext that the subject had given his ‘consent’. Such practices should be forbidden even if undertaken with the consent of the subject. In reply, it was argued that consent given under pressure could never be regarded as ‘free’ consent. It was unthinkable that anyone would freely submit himself to torture or cruel, inhuman or degrading practices.
Dinstein and Nowak disagreed on this aspect. According to Dinstein, ‘torture, unlike medical or scientific experimentation, is unlawful even if the victim (say, for masochistic reasons) agreed to be subjected to it’.Footnote 73 According to Nowak, ‘both the wording of the provisions and the travaux préparatoires tend to indicate the contrary’.Footnote 74 The 2001 Concluding Observations of the HRC on the Netherlands nevertheless brought some clarification. At the time, Section 3 of the Dutch Medical Research Involving Human Subjects Act (WMO) mentioned that research protocols could be approved only if, among other conditions, ‘it is reasonable to expect that the risk to and burden for the subject will be in proportion to the potential value of the research’.Footnote 75 However, the HRC affirmed that this condition did not comply with Article 7 of the ICCPR. In fact, the Committee was ‘concerned at the general criterion whereby proportionality is assessed by balancing the risks of the research to the subject against the probable value of the research’.Footnote 76 It considered ‘that this rather subjective criterion must be qualified by a limitation beyond which the risks are so great to the individual that no measure of expected benefit can outweigh them’.Footnote 77 This limitation is that of ‘severe risks’.Footnote 78 It thus appears that, whatever is the goal of the experimentation (non-therapeutic or therapeutic), research in general should not come with severe risks. It must be acknowledged that the observance of a ‘red line’ is obvious in the legislation of some states, where non-therapeutic research (that is, research deprived of any direct benefit to the subject) is contingent upon minimal riskFootnote 79 or the absence of undue risk.Footnote 80
The question then becomes the following: should the experimentation stop when the expected benefit for the subject is great, but comes with significant risk? The practice of the HRC suggests that such research would still be lawful. In fact, the Netherlands reacted to the 2001 Concluding Observations, emphasising in the Fourth Periodic Report that in some situations such limit – that of ‘severe risk’ – was not desirable.Footnote 81 The Dutch government argued, for instance, that potential lasting damage caused by medical research could be justified when subjects suffered from a deadly disease.Footnote 82 This interpretation was not challenged by the Committee in the 2009 Concluding Observations.Footnote 83 In addition, there was a dispute regarding Section 3a of the WMO, which concerned persons under the age of 18 and those ‘who cannot be deemed capable of reasonably assessing their interests in the matter’.Footnote 84 It prohibited experimentation on such individuals unless the research was ‘of direct benefit to the subjects’ (therapeutic research) or ‘could not be conducted without the participation of persons of the same category as the subject, provided that the risk associated with participation is negligible and the burden minimal’ (non-therapeutic research).Footnote 85 The HRC insisted, in the 2001Footnote 86 and the 2009Footnote 87 Concluding Observations, that minors and incapacitated persons should not be subjected to medical experiments ‘which do not directly benefit’ them. In contrast, the possibility of carrying out therapeutic research on those subjects was not challenged by the Committee either. The Netherlands, in the Fifth Periodic Report, underlined that clinical trials on minors could be conducted only where ‘some benefit for the population represented by the minor concerned’ was expected and where ‘a clinical trial will pose only minimal risk to, and will impose minimal burden on, the minor concerned in comparison with the standard treatment of the minor's condition’.Footnote 88 This means that – in the absence of a direct benefit for the subject – experimentation is acceptable only when the minor suffers from a condition, receives treatment for that condition, and the risk/burden incurred by the minor is minimal compared with this treatment.Footnote 89 It must be noted that the Dutch legislation applies similar conditions to incapacitated persons.Footnote 90 In addition, Article 3(1)(c) of the WMO has set a new general requirement, according to which the interest of the test subject and future patients should be proportional to the risk incurred by the subject. None of these aspects were questioned by the HRC in the 2019 Concluding Observations.Footnote 91
What conclusions can be drawn with regard to experimentation in general, and BCIs in particular? When it comes to experimentation in general, non-therapeutic research (that is, research with no direct benefit for the subject) is lawful provided that the person is not subjected to ‘severe risk’. Non-therapeutic research is generally prohibited on minors and incapacitated persons.Footnote 92 However, it would be acceptable for minors (and perhaps incapacitated persons) in situations where the experimentation may present a benefit for the population they represent, and where it comes only with minimal risk and burden (in comparison with the standard treatment received by the subject for the condition). In contrast, therapeutic research (that is, research with potential direct benefit for the subject) is permissible on any subject, provided that the proportionality between risk and benefit is respected. It is worth underlining that experimentation carried out with a view to enhancing the capacities of the subject may be described as non-therapeutic, and this legal framework obviously influences the lawfulness of experimentation involving BCIs. This means that healthy persons who cannot expect any direct benefit from the research may consent only to test non-invasive devices. The latter are less risky because they ‘rest outside the brain and try to capture brain signals’,Footnote 93 usually through scalp recording (EEG),Footnote 94 as well as through other external devices, such as magnetoencephalography (MEG) or functional magnetic resonance imaging (fMRI). Such an interpretation is confirmed by a study of the research protocols authorised on healthy volunteers in France,Footnote 95 Germany,Footnote 96 Singapore,Footnote 97 SwitzerlandFootnote 98 and the US.Footnote 99 In fact, they involve only non-invasive devices, even when the experimentation aims to develop a therapeutic device.Footnote 100 Research protocols in France, Germany, Greece, Portugal, Singapore and the US also confirm that experimentation with BCIs on minors are authorised only when the purpose is primarily therapeutic,Footnote 101 while psychiatric pathologies or mental retardation remain a factor of exclusion for most trials.Footnote 102 In contrast, persons who may benefit from the development of BCIs may consent to higher risk, and be subjected to invasive and partially invasive devices. As once underlined by Konrad and Shanks:Footnote 103
Clinicians who are involved in present day neurological device implants weigh this risk in the context of clinical loss, and this may well be acceptable to the patient and clinician when considering BCI implants in patients who are already paralyzed or blind when considering the potential benefit of restoring lost function.
This interpretation is also confirmed by the review of existing research protocols. In fact, several clinical trials intend to allow paralysed persons to moveFootnote 104 and/or communicateFootnote 105 again. Some research protocols involving paralysed persons, however, are riskier than others. A study in the US, for instance, involved the implantation of sensors over the motor cortex of participantsFootnote 106 (that is, on the brain surface); another study relied on the implantation of arrays in the primary motor cortex and the dominant brain hemisphere of the subjects.Footnote 107 For the moment, most of the trials described above (as well as further tests carried out to prevent cognitive decline in elderly people,Footnote 108 visual loss,Footnote 109 autism,Footnote 110 depression,Footnote 111 or pain managementFootnote 112) were carried out with a view to find a ‘cure’. However, what happens with devices that may have the potential for both therapy and enhancement,Footnote 113 or even for the sole purpose of enhancement? For instance, the objective of the DARPA Human Assisted Neural Devices programme has always been to ‘develop the scientific foundation for novel concepts that will improve performance on the battlefield as well as technologies for enhancing the quality of life of paralyzed veterans’.Footnote 114 In this framework the subject, a paralysed woman, agreed to have two sensors implanted on the surface of her brain in the area responsible for hand and arm movement.Footnote 115 At the beginning of the experiment she performed simple tasks such as controlling a robotic arm, feeding herself, and giving a high five and thumbs up.Footnote 116 Over the years, she successively managed to pilot, on a flight simulator, a Cessna, a F-35 fighter and three fighters at once. Here again, proportionality would apply. Experimentation that may have direct benefit for the subject would be considered to be therapeutic and the subject may consent to higher risk (such as the implantation of captors on the brain). In contrast, experimentation with no expected direct benefit would be considered non-therapeutic and participants may not be subjected to severe risk. It may be emphasised, though, that this framework does not prevent the use, for the purpose of enhancement, of devices that were originally developed as therapeutic. As once underlined by Juengst and Moseley:Footnote 117
[N]o matter how the line is drawn, most biotechnological interventions that could be seen as problematic if used as enhancements will not need to be justified as enhancements in order to be developed and approved for clinical use. This is because most such interventions will also have legitimate therapeutic applications. Indeed, most biomedical tools with potential for enhancement uses will first emerge as therapeutic agents.
3.2.2. The Council of Europe Framework
While the ECtHR has adopted a clear position on the meaning of free consent, nothing similar exists with regard to the evaluation of risk and benefit. It is difficult to determine clearly whether a healthy subject could consent to undergo risky experiments and whether such treatment might be contrary to Article 3 of the ECHR. In fact, there seems to be a gap in the ECHR, and attempts have been made to fill it. In a non-binding recommendation the Committee of Ministers of the Council of Europe proposed that the governments of member states adopt legislation in conformity with several principles. Among these is the following: ‘[t]he risks incurred by a person undergoing medical research must be kept to a minimum. The risks should not be disproportionate to the benefits to that person or the importance of the aims pursued by the research’.Footnote 118 This recommendation directly inspired the Oviedo Convention,Footnote 119 which therefore requires that ‘the risks which may be incurred by that person are not disproportionate to the potential benefits of the research’.Footnote 120 Article 6(2) of the most recent Additional Protocol to the Convention on Human Rights and Biomedicine, concerning Biomedical Research mentions:Footnote 121
[W]here the research does not have the potential to produce results of direct benefit to the health of the research participant, such research may only be undertaken if the research entails no more than acceptable risk and acceptable burden for the research participant.
The Additional Protocol has only been ratified by 12 members of the Council of Europe.Footnote 122 It is arguable that it limits the scope for experimentation. First, and apparently in contrast with the recommendation of the Committee of Ministers and the provisions of the Oviedo Convention, the risk cannot be balanced only against the importance of the research. Research is permissible only in two situations: (i) when a potential and direct benefit for the research participant is present or (ii) in the absence of such potential direct benefit, when the experimentation is not accompanied by unacceptable risk and burden. Second, the ‘acceptable risk and acceptable burden’ seem more restrictive than the HRC limitation of ‘severe risk’.Footnote 123 One can postulate that the participation of healthy volunteers will probably be harder to allow under this protocol. Research on ‘persons not able to consent’ was also limited to two situations by the Oviedo Convention and the Additional Protocol concerning biomedical research.Footnote 124 First, research is permissible when the results ‘have the potential to produce real and direct benefit’ for the ‘health’ of the participant, and that ‘research of comparable effectiveness cannot be carried out on individuals capable of giving consent’.Footnote 125 Second, and lacking ‘the potential to produce results of direct benefit to the health’ of the subject, research is ‘exceptionally’ permissible when it may deliver ‘results capable of conferring benefit to the person concerned or to other persons in the same age category or afflicted with the same disease or disorder or having the same condition’.Footnote 126 In this case, minimal risk and burden are required.Footnote 127
What conclusions can be drawn with regard to BCIs? The Oviedo Convention is less restrictive than its Additional Protocol concerning biomedical research. In fact, under the Oviedo Convention, the admissible risk must be proportionate to the potential benefits of the research. This means that the prospects for progress in the medical sector, or even for the military, might justify higher risk for the subject, at least if the subject is able to consent. For instance, it may be permissible to put electrodes over the brain of a participant even if no personal benefit is expected. In contrast, this would never be acceptable under the Additional Protocol, to the extent that such implantation would probably go beyond acceptable risk and burden. Again, those who might benefit from the research, such as paralysed persons, may consent to try partially invasive or invasive devices even if they come with higher risk, whereas healthy volunteers may only take part in experimentation that involves non-invasive BCIs. Research on persons who are unable to consent would, under both instruments, be restricted. That being said, it is necessary to determine whether and how those general rules would apply to experimentation carried out on members of the armed forces. In contrast with civilians, they are actually subject to the orders of the military commander, they must protect the country at the risk of their own lives, and some of their rights may be restricted.
4. The Application of Free Consent and Proportionality to the Experimental Enhancement of Military Forces
As mentioned in the introduction, there is growing interest from armies for BCIs, which wish to use them for the purpose of augmenting the capacities of soldiers. However, neither the framework of the ICCPR nor that of the Council of Europe has expressly tackled the question of experimentation carried out on soldiers. Members of the armed forces find themselves in a very specific situation. As a result of military hierarchy they are indeed particularly vulnerable to coercion, and are assigned with a mission of a unique nature: to protect national security, sometimes at the risk of their lives. Free consent, however, must be strictly respected (4.1), whereas proportionality may be adapted to this special context in certain situations (4.2).
4.1. The Strict Respect for Free Consent
Over the last decade or so, instances of abusive experiments into which soldiers had been coerced have come to light,Footnote 128 and a recent example is that of the experimental anthrax vaccines in the 1990s. Indeed, it was revealed that American, Australian, British and Israeli soldiers had been put under pressure to receive the injection.Footnote 129 In fact, this illustrates an inherent tension in the army – reconciling free consent and superior orders – and raises the question whether soldiers are in a position to refuse to participate.Footnote 130 In theory, free consent of the members of the armed forces must be respected, both under the ICCPR and the ECHR. It thus appears that even if the consent of soldiers is not necessary for them to be sent onto the battlefield, it is required in the framework of experimentation.Footnote 131
Article 7 of the ICCPR indeed makes clear that ‘no one shall be subjected without his free consent to medical or scientific experimentation’. The HRC also mentioned that persons ‘not capable of giving valid consent’ should ‘not be subjected to any medical or scientific experimentation that may be detrimental to their health’.Footnote 132 If some states – which include AustraliaFootnote 133 and South AfricaFootnote 134 – consider soldiers to be vulnerable subjects,Footnote 135 this view has not been confirmed by the HRC for the moment.Footnote 136 However, the lawfulness of the US legislationFootnote 137 was indeed questioned by the Committee. It noticed that the US President was authorised, in some situations, ‘to waive the prior informed-consent requirement for the administration of an investigational new drug to a member of the US Armed Forces’.Footnote 138 This was possible ‘if the President determines that obtaining consent is not feasible, is contrary to the best interests of the military members, or is not in the interests of U.S. national security’.Footnote 139 On this occasion the Committee highlighted that the US ‘should ensure that it meets its obligation under article 7 of the Covenant not to subject anyone without his/her free consent to medical or scientific experimentation’, then underlined the non-derogable character of Article 7.Footnote 140 It may be noted, though, that the US legislation has remained unchangedFootnote 141 and that nothing was said by the HRC about such provision in the 2014 Concluding Observations.Footnote 142
As for the ECHR, it was made clear that members of the armed forces are not deprived of their fundamental freedoms, even if some limits are acceptable for the purpose of military discipline. The specificity of their situations is thus expressly established. This is how the ECtHR acknowledged in 1976 that the Convention applies in principle to members of the armed forces and not only to civilians,Footnote 143 and ‘does not stop at the gates of army barracks’.Footnote 144 However, it was also found that in interpreting and applying the rules of the ECHR, the Court must bear in mind the particular characteristics of military life and its effects on the situation of individual members of the armed forces.Footnote 145 Restrictions on certain freedoms, in particular, are permitted ‘where there is a real threat to military discipline’Footnote 146 or the ‘operational effectiveness’ of the armed forces,Footnote 147 provided that such risk is ‘substantiated by specific examples’.Footnote 148 Moreover, the rights protected by the Convention do not cease in order to apply with operational needs, or even the deployment of troops on the battlefield. The ECtHR indeed underlined that states have a ‘primary duty’, which consists of implementing ‘rules geared to the level of risk to life or limb’ that may result ‘from the nature of military activities and operations’, as well as ‘from the human element that comes into play when a State decides to call up ordinary citizens to perform military service’.Footnote 149 In addition, the Committee of Ministers emphasised that ‘[m]embers of the armed forces should not be exposed to situations where their lives would be avoidably put at risk without a clear and legitimate military purpose or in circumstances where the threat to life has been disregarded’.Footnote 150 As underlined above, the ECtHR found in VC v Slovakia that carrying out a medical procedure without the informed consent of a mentally competent adult was contrary to Article 3 of the ECHR. A particularity of Article 3 – according to which ‘[n]o one shall be subjected to torture or to inhuman or degrading treatment or punishment’ – is that it does not tolerate any restriction. This means that the requirement of free consent must be obtained and strictly observed before submitting soldiers for experimentation.Footnote 151
4.2. The Possible Adaptation of Proportionality
In the context of experimentations with members of the armed forces the proportionality test requires specific attention.Footnote 152 It is indeed obvious that in general soldiers do not have a direct expected benefit in taking part in experiments. However, two questions are of importance. First, could members of the armed forces participate in trials that may have only hypothetical future benefit? Second, could they take part in experiments that have benefit only for other battalions or national security?
With regard to the first question, it must be underlined that it might be tempting for members of the armed forces to take part in risky experiments. As underlined by some commentators, ‘the instinct for self-preservation is likely to lead soldiers to grasp at any means of improving their chances of surviving battle, including exposing themselves to risks in order to gain access to experimental enhancements’.Footnote 153 In fact – and if free consent is essential to prevent coercion stemming from military hierarchy – proportionality might then prove to be essential in order to protect soldiers against themselves. The use of BCIs has huge potential in decreasing combat-related risk for war fighters. Telepresence, for instance, would allow the remote control of a robot or a drone based on brainwaves and might prevent soldiers from having to be physically present on the battlefield. BCIs may also be used in modern threat-detection systems,Footnote 154 exoskeletons, and allow ‘silent talk’ – that is, ‘user-to-user communication on the battlefield without the use of vocalised speech through analysis of neural signals’.Footnote 155 In a nutshell, BCIs would reduce the risk of soldiers being exposed to fire and of being detected on the battlefield; it would improve communication between them, and give them more strength and endurance.
In addressing the second question, it must be noted that some authors clearly take a position in favour of a re-evaluation of the proportionality test, arguing that soldiers should be allowed ‘to trade risk on the battlefield with risk in the laboratory’.Footnote 156 Mehlman and Li, for instance, consider that soldiers ‘are obliged to obey lawful orders’, and that in the army there is ‘an obligation to promote the welfare of subordinates to the extent consistent with the welfare of the unit, the accomplishment of the mission, and the safeguarding of the state’.Footnote 157 They suggest replacement in the army of the ‘civilian principle of beneficence’ with ‘the principle of proportionality’.Footnote 158 This means that imposing a biomedical risk on a soldier is acceptable only when there is no less risky alternative to accomplish a legitimate military objective, and the nature and degree of the risk are outweighed by the military advantage sought to be gained.Footnote 159 They consider that a research risk that would be deemed excessive for civilian subjects might yet be deemed acceptable for military subjects ‘under the principle of proportionality if the military objectives were sufficiently important’.Footnote 160 Hence, ‘while the risks might be greater than for civilians, so might be the corresponding benefits to the unit, mission, or country’.Footnote 161
In fact, the lawfulness of such experimentation depends on the legal framework that is considered. It is clear that most experimental BCIs bring no personal benefit for the healthy soldier – perhaps, at best, one only of a hypothetical future nature. As previously underlined, the HRC interpretation is that under Article 7 of the ICCPR, the subject should never be exposed to ‘severe risks’. What counts is that the ‘red line’ is never crossed. Testing an invasive BCI on a healthy soldier is certainly not acceptable in this context as manipulating the brain could result in severe paralysis, mental disorder or death. In contrast, non-invasive devices hardly constitute a risk for the subject, and testing them on soldiers would not be a problem. The case of partially invasive devices is a trickier question. In fact, ‘[t]he partially invasive BCIs have less exposure of scar tissue formation as compared to invasive BCI … ECoG does not damage any neurons because no electrodes penetrate the brain’.Footnote 162 However, they still involve skull surgery,Footnote 163 and one can argue that such a procedure would not comply with ‘severe risk’. The so-called ‘minutely invasive’ techniques currently being developed by DARPA in the framework of the Next-Generation Nonsurgical Neurotechnology programme (N3) would, for their parts, probably comply with ‘severe risk’. The point with such techniques is to ‘permit nonsurgical delivery of a nanotransducer’.Footnote 164 Even though the nanotransducer would be directly in contact with neurons, ‘opening’ the head of the subject would not be required in this situation.
While the evaluation of proportionality in light of the ECHR remains obscure, it was also mentioned previously that there are two other instruments on the European scene: the Oviedo Convention and its Additional Protocol, concerning biomedical research. Their approaches to proportionality differ. The Convention states that ‘the risks which may be incurred by that person are not disproportionate to the potential benefits of the research’.Footnote 165 The Protocol, in contrast, states that ‘where the research does not have the potential to produce results of direct benefit to the health of the research participant, such research may only be undertaken if the research entails no more than acceptable risk and acceptable burden’ for the participant.Footnote 166 If a literal interpretation of the Oviedo Convention theoretically supports the possibility of putting soldiers’ lives at risk in order to develop military BCIs for the benefit of future battles, other battalions or national security, as they would obviously have huge ‘potential benefits’, most states would in practice be restricted by their commitment to comply with the ICCPR, which imposes the limitation of ‘severe risk’. As for the Additional Protocol, the criteria of ‘acceptable risk’ and ‘acceptable burden’ would never tolerate experimentation with invasive devices, nor probably with partially invasive devices. However, the testing of ‘minutely invasive’ techniques, as foreseen in the US, might comply with these conditions. A potential benefit for national security or future battles might thus justify a higher risk, but not any risk.
5. Conclusion
The emergence of enhancement technologies challenges our assumptions about biomedical research: medical progress may not only cure but could also augment our capacities. Indeed, it appears that human subjects may take part in experimentation that is deprived of personal therapeutic interest, or only partially justified by it. Free consent and proportionality, which were originally present in non-binding instruments, were finally given legal consecration by the ICCPR and the HRC. Under this regime consent must be obtained, whether directly or by proxy. Persons who are not able to give free consent may not take part in research ‘that may be detrimental to their health’ or ‘do[es] not directly benefit’ them, while ‘the only experimental treatment compatible with article 7 would be treatment chosen as the most appropriate to meet the medical needs of the individual’.Footnote 167 The participation of other volunteers is acceptable, as long as they are not subject to ‘severe risk’. It thus means that participants who are unable to consent freely may be subject to experimentation only with therapeutic BCIs (or those of a primarily therapeutic nature), and when they or, under certain conditions, persons of the same category could benefit from such a process. At the opposite end, healthy participants could agree to test therapeutic and enhancement devices, as long as they do not undergo severe risks, thus precluding invasive and partially invasive BCIs.
If the requirement of free consent was also acknowledged by the ECtHR on the basis of Article 3 of the ECHR, no approach to proportionality was made. The Oviedo Convention and its Additional Protocol concerning biomedical research have specific provisions on this issue, but their respective approaches differ. Under the Convention, the risks should not be disproportionate compared with the ‘potential benefits of the research’. Under the Additional Protocol, the volunteer should not be subjected to more than ‘acceptable risk and acceptable burden’ when the research project has no expected benefit for the subject.
With regard to members of the armed forces, it appears that free consent must still be obtained under both the ICCPR and the ECHR. The limit of ‘severe risk’ defined by the HRC also applies to such persons and prevents experimentation with invasive or partially invasive devices. The Oviedo Convention brings no further limitations, as military enhancement is a type of research that has great ‘potential benefits’ for future battles, national security and other war fighters. The Additional Protocol confirms that invasive or partially invasive devices should probably not be tested on soldiers, as they come with unacceptable risk for the subject. As for international humanitarian law, it appears that prisoners of war and protected persons may consent only to be subjected to experiments required for their state of health or medical treatment. Experimentation with BCIs is thus conditioned upon a personal therapeutic benefit.
Discussions of the regulation of human enhancement will probably continue, both among international lawyers and at the domestic level. However, international agreement on this issue is an unlikely outcome on a short-term basis. If it requires time to allow a better understanding of the technology at stake, it also means that – in the absence of intervention from judicial or quasi-judicial bodies – some research-related issues are unlikely to receive due attention in the near future. For instance, a convergence of the views of human rights instruments on proportionality would be desirable, as well as further clarification regarding the room for manoeuvre that states and labs actually have within the HRC's red line of ‘severe risks’. In addition, states do not play by the same rules, which may be particularly unfair in some ways. China, for instance, has not ratified the ICCPR but strives to become a leader in the field of military BCIs.Footnote 168 This means that, assuming that China does not have satisfactory legal protection for human experiments,Footnote 169 research could progress more quickly there, as some projects which would be prohibited by international law in certain parts of the world could still be carried out in the ‘Middle Kingdom’. Technologies of western armies could thus be surpassed by those of countries that are less respectful of human rights. Another problem might be the ‘subcontracting’ of research. To keep pace with others, laboratories might be inclined to relocate some experiments in states that have little respect for human rights. A step back, for the time being, is excluded. As well synthesised by the French Secretariat-General for National Defence and Security (SGDSN):Footnote 170
The current situation is that states that play an active role in [the development of] neurosciences are not ready to embark on a regulation process of their military applications. The US considers that [America's] advance [in new technologies] allows them to gain durable capacity advantages, which will allow them to enhance the superiority of their forces and their power. In a nutshell, the situation of neurosciences’ military applications [is comparable with] the situation that existed in the nuclear field during the sixties, and before the negotiations about the NPT started: a context of non-cooperation and competition. Each party is expecting to overtake the others, and has technical reasons to assume it.
