{"id":87602,"date":"2026-11-17T19:51:13","date_gmt":"2026-11-17T19:51:13","guid":{"rendered":"https:\/\/www.europesays.com\/conflicts\/87602\/"},"modified":"2026-11-17T19:51:13","modified_gmt":"2026-11-17T19:51:13","slug":"the-law-on-brain-technologies-in-warfare-series-the-brain-firing-lawfulness-and-liability-of-weapon-control-through-bci","status":"publish","type":"post","link":"https:\/\/www.europesays.com\/conflicts\/87602\/","title":{"rendered":"The Law on Brain Technologies in Warfare Series \u2013 The Brain Firing: Lawfulness and Liability of Weapon Control through BCI"},"content":{"rendered":"<p>Editors\u2019 note: This post is an entry in our series on Air Commodore (ret.) William H. Boothby\u2019s book, The Law on Brain Technologies in Warfare <a href=\"https:\/\/www.e-elgar.com\/shop\/usd\/the-law-on-brain-technologies-in-warfare-9781035394449.html\" rel=\"nofollow noopener\" target=\"_blank\">published<\/a> by Elgar International.<\/p>\n<p>Soldiers may soon direct weapon systems using only their thoughts. Brain-computer interfaces (BCIs) can connect a person\u2019s brain to a computing device. The computer picks up neural signals generated by the brain and translates them into a command for a physical gadget, allowing the user to mentally control that gadget. Employing BCIs to direct weapons has already been <a href=\"https:\/\/www.darpa.mil\/news\/2019\/nonsurgical-brain-machine-interfaces\" rel=\"nofollow noopener\" target=\"_blank\">researched<\/a> and has even earned a <a href=\"https:\/\/www.freepatentsonline.com\/y2005\/0090756.html\" rel=\"nofollow noopener\" target=\"_blank\">patent<\/a>.<\/p>\n<p>After sketching possible BCI-enabled set-ups for weapon control, this post highlights legal challenges the technology poses to the law of armed conflict and to international criminal law.<\/p>\n<p>Weapon Control through BCIs<\/p>\n<p>Various scenarios for weapon control are conceivable: a commander may remotely direct a ground vehicle <a href=\"https:\/\/international-review.icrc.org\/articles\/merging-man-and-machine-a-legal-assessment-of-brain-computer-interfaces-in-armed-conflict-928\" rel=\"nofollow noopener\" target=\"_blank\">equipped<\/a> with a gun, a weaponized drone, or a missile firing mechanism aboard a vessel, to name just a few. In principle, there are two types of BCIs: active and reactive.<\/p>\n<p>Active BCIs require the soldier to focus on giving a certain command. For example, the computer may recognize the neural signals evoked when the soldier imagines pressing a button with his hand. Thereupon, the computer transmits the instruction to fire to the connected weapon. This manner of control demands an intense, conscious <a href=\"https:\/\/doi.org\/10.1080\/21507740.2019.1704918\" rel=\"nofollow noopener\" target=\"_blank\">effort<\/a>.<\/p>\n<p>Reactive BCIs decipher neural signals generated subconsciously, before the operator becomes aware of them. Imagine that you accidentally place your hand on a hot stove top. Your brain will elicit signals to immediately retract it. These signals are an instantaneous, subconscious reaction. For military BCIs, such signals can be provoked by rapid serial visual presentation (RSVP). The soldier, tasked with spotting a target, is shown <a href=\"https:\/\/ieeexplore.ieee.org\/document\/9390035\" rel=\"nofollow noopener\" target=\"_blank\">images<\/a> of the battlefield at very short intervals. Upon recognizing the target, the soldier\u2019s brain emits a subconscious neural response, which the BCI then decodes and transmits to the weapon.<\/p>\n<p>Weapon control is arguably one of BCI\u2019s most promising applications for militaries. It permits remote command without depending on a physical control station. Reactive BCIs, on top, enable ultrarapid decisions. Target engagement based on subconscious neural signals is possible in 250 to 500 milliseconds after visual perception, which is the <a href=\"https:\/\/academic.oup.com\/lril\/article-abstract\/2\/2\/201\/944587?redirectedFrom=fulltext\" rel=\"nofollow noopener\" target=\"_blank\">fastest<\/a> method currently imaginable to operate any device.<\/p>\n<p>International Humanitarian Law<\/p>\n<p>Because thought-controlled weapons are unprecedented, I first address their general legality under weapons law and then turn to relevant aspects of targeting law.<\/p>\n<p>Weapons Law <\/p>\n<p>Initially, the question arises whether such a system can be legally used in armed conflict under any circumstance. Considering the man-machine collaboration constitutes a <a href=\"https:\/\/international-review.icrc.org\/articles\/merging-man-and-machine-a-legal-assessment-of-brain-computer-interfaces-in-armed-conflict-928\" rel=\"nofollow noopener\" target=\"_blank\">weapon system<\/a>, i.e. a means of warfare, a review under <a href=\"https:\/\/ihl-databases.icrc.org\/en\/ihl-treaties\/api-1977\/article-36\" rel=\"nofollow noopener\" target=\"_blank\">Article 36<\/a> of Additional Protocol I (AP I) (considered <a href=\"https:\/\/doi.org\/10.1093\/law\/9780198855309.003.0011\" rel=\"nofollow noopener\" target=\"_blank\">customary international law<\/a>) is in order. Accordingly, the BCI must be capable of complying with international humanitarian law a priori,\u00a0independent of a specific operational context. Closer scrutiny is demanded to determine whether the technology is <a href=\"https:\/\/ihl-databases.icrc.org\/en\/ihl-treaties\/api-1977\/article-48?activeTab=\" rel=\"nofollow noopener\" target=\"_blank\">inherently indiscriminate<\/a> or incapable of ensuring <a href=\"https:\/\/ihl-databases.icrc.org\/en\/ihl-treaties\/api-1977\/article-51?activeTab=#:~:text=(b)%20an%20attack%20which%20may%20be%20expected%20to%20cause%20incidental%20loss%20of%20civilian%20life%2C%20injury%20to%20civilians%2C%20damage%20to%20civilian%20objects%2C%20or%20a%20combination%20thereof%2C%20which%20would%20be%20excessive%20in%20relation%20to%20the%20concrete%20and%20direct%20military%20advantage%20anticipated.\" rel=\"nofollow noopener\" target=\"_blank\">proportionate<\/a> targeting <a href=\"https:\/\/ihl-databases.icrc.org\/en\/ihl-treaties\/api-1977\/article-57?activeTab=#:~:text=%C2%A0%20(iii)%20refrain%20from%20deciding%20to%20launch%20any%20attack%20which%20may%20be%20expected%20to%20cause%20incidental%20loss%20of%20civilian%20life%2C%20injury%20to%20civilians%2C%20damage%20to%20civilian%20objects%2C%20or%20a%20combination%20thereof%2C%20which%20would%20be%20excessive%20in%20relation%20to%20the%20concrete%20and%20direct%20military%20advantage%20anticipated%3B\" rel=\"nofollow noopener\" target=\"_blank\">decisions<\/a>.<\/p>\n<p>The active BCI lets the operator consciously select targets much like a traditional remote-controlled system. To effectively discriminate between targets, the system must reliably respond to the user\u2019s commands. Particularly, the computer must differentiate between mere ideas (\u201c<a href=\"https:\/\/doi.org\/10.1080\/21507740.2019.1704918\" rel=\"nofollow noopener\" target=\"_blank\">hazy intentions<\/a>\u201d) and the final resolution to engage. Even when the soldier is under immense stress, anxiety, or in pain, the correct intention ought to be identified. Provided comprehensive testing can ensure reliable functioning, the active BCI does not pose issues of legality.<\/p>\n<p>The reactive BCI, in contrast, does not allow the soldier to consciously reflect or even to veto a targeting decision. Before the operator is aware of subconsciously identifying a target, the weapon will have struck it. Thus, the soldier lacks conscious control over the \u201c<a href=\"https:\/\/international-review.icrc.org\/articles\/merging-man-and-machine-a-legal-assessment-of-brain-computer-interfaces-in-armed-conflict-928\" rel=\"nofollow noopener\" target=\"_blank\">moment of command<\/a>.\u201d This makes it hard to predict whether the user will distinguish lawful from unlawful targets as well as conduct a proper\u2014subconscious\u2014proportionality assessment, where necessary.<\/p>\n<p>Having a human operator at the core of the targeting process, while simultaneously lacking control over decisions made, constitutes an unusually challenging set-up. It challenges predictability. With conventional weapons, the predictability of a targeting decision is ordinarily uncontroversial. As long as a manually operated weapon like a hand-held gun can be directed against a specific target, it is assumed the target selection is discriminate. Once certified that the gun strikes the envisaged target, the outcome of its use is sufficiently predictable. This is because the shooter controls the target selection process by aiming the gun. Differently put, a gun per se respects the principle of distinction because its targeting process is fully controlled by the soldier using it.<\/p>\n<p>The reactive BCI cannot grant this amount of control. The underlying assumption of generating a predictable outcome fails. This compels a positive test for the system\u2019s overall adherence to distinction and proportionality. Generally, where human supervision over the targeting process is absent, it is necessary to assess under Article 36 of AP I whether the weapon system can <a href=\"https:\/\/www.elgaronline.com\/monochap\/book\/9781035394456\/chapter4.xml\" rel=\"nofollow noopener\" target=\"_blank\">comply<\/a> with the targeting principles.<\/p>\n<p>One such test may be to ascertain how much conscious human control is yet present in the targeting process. This notion ensures that the soldier\u2019s intention to strike lawful targets actually manifests in predictably striking lawful targets. There appears to be no definitive threshold of how much control, and thus how much predictability, is prescribed.<\/p>\n<p>Similar to the discussion on lethal autonomous weapon systems, one could argue that human oversight at the stage of engineering and testing suffices to engender predictable results. Still, even with extensive testing in a simulation, a subconscious decision may be premature. Possibly, it would have been corrected had there been just one more conscious moment of deliberation.<\/p>\n<p>I have discussed this issue in more depth <a href=\"https:\/\/international-review.icrc.org\/articles\/merging-man-and-machine-a-legal-assessment-of-brain-computer-interfaces-in-armed-conflict-928\" rel=\"nofollow noopener\" target=\"_blank\">elsewhere<\/a>. As a final note for this post, one must keep in mind that the legality of BCI systems will highly depend on the employment scenario(s) for which they are built. The review is substantially altered, for example, in case the reactive BCI\u2019s \u201cnormal or expected <a href=\"https:\/\/ihl-databases.icrc.org\/en\/ihl-treaties\/api-1977\/article-36\/commentary\/1987?activeTab=\" rel=\"nofollow noopener\" target=\"_blank\">circumstances<\/a>\u201d of use are operations not endangering civilian targets whatsoever.<\/p>\n<p>Targeting Law<\/p>\n<p>Assuming one considers both BCI-controlled weapon systems permissible at all, they ought to obey the targeting principles, which are also customary in nature. Particularly relevant are distinction and proportionality (in the context of the specific attack) as well as the duty to take <a href=\"https:\/\/ihl-databases.icrc.org\/en\/ihl-treaties\/api-1977\/article-57?activeTab=\" rel=\"nofollow noopener\" target=\"_blank\">precautions<\/a>.<\/p>\n<p>Regarding the active BCI, its specific technicalities must be taken into account. BCIs may not only transmit the soldier\u2019s thoughts but also supply input from the controlled device (so-called <a href=\"https:\/\/link.springer.com\/chapter\/10.1007\/978-3-030-72254-8_20\" rel=\"nofollow noopener\" target=\"_blank\">bi-directional BCIs<\/a>). Detailed information from the battlefield could be transmitted directly to the user\u2019s brain. This arguably influences a given target assessment. Alternatively, bi-directional BCIs permit <a href=\"https:\/\/doi.org\/10.1017\/pls.2022.2\" rel=\"nofollow noopener\" target=\"_blank\">altering<\/a> the soldier\u2019s mood. Such a modified state of mind could distort the operator\u2019s judgment, leading to less precise distinction between, say, fighters of armed groups from civilians. Further, the soldier might adversely weigh the military necessity against incidental loss of civilian life.<\/p>\n<p>The reactive BCI additionally raises concerns regarding the duty to take all feasible precautions in the choice of means and methods of attack with a view to avoiding, and in any event to <a href=\"https:\/\/ihl-databases.icrc.org\/en\/ihl-treaties\/api-1977\/article-57?activeTab=#:~:text=%C2%A0%20(ii)%20take%20all%20feasible%20precautions%20in%20the%20choice%20of%20means%20and%20methods%20of%20attack%20with%20a%20view%20to%20avoiding%2C%20and%20in%20any%20event%20to%20minimizing%2C%20incidental%20loss%20of%20civilian%20life%2C%20injury%20to%20civilians%20and%20damage%20to%20civilian%20objects%3B\" rel=\"nofollow noopener\" target=\"_blank\">minimizing<\/a>, incidental loss of civilian life. Given the subconscious nature of targeting, the reactive BCI cannot, by design, constitute a choice of means that minimizes collateral harm compared to conscious decision-making. Conscious decision-making (as in active BCIs) will inherently produce more, or at least equally, discriminate and proportionate targeting decisions.<\/p>\n<p>Hence, a feasible precaution is to ensure the soldier makes a conscious targeting decision. This naturally limits the employment of reactive BCIs to scenarios in which either striking civilian targets is precluded per se, or the military advantage of increased speed is sweeping once the attack has been decided upon.<\/p>\n<p>International Criminal Law <\/p>\n<p>BCIs may be used to commit international core crimes. To hold the operator accountable under international criminal law, the mentally-initiated strike of a weapon must constitute an actus reus, and the soldier must satisfy the mens rea. Both seem problematic. Subsequently, I map out the most elementary issues based on the <a href=\"https:\/\/www.icc-cpi.int\/sites\/default\/files\/2024-05\/Rome-Statute-eng.pdf\" rel=\"nofollow noopener\" target=\"_blank\">Rome Statute<\/a>.<\/p>\n<p>Actus Reus<\/p>\n<p>The actus reus (material element of the crime) refers to the action the perpetrator must exert to carry out a crime under the Rome Statute. At its core, the soldier must engage in specified conduct, or perform an <a href=\"https:\/\/academic.oup.com\/oxford-law-pro\/book\/57215\/chapter\/474133420\" rel=\"nofollow noopener\" target=\"_blank\">act<\/a>.<\/p>\n<p>What such an act denotes is not defined in the Rome Statute. Frequently referred to in this context is the <a href=\"https:\/\/www.legal-tools.org\/doc\/08d77d\/pdf\" rel=\"nofollow noopener\" target=\"_blank\">U.S. Model Penal Code<\/a> describing an act or action as \u201ca bodily movement.\u201d Because the mental command of a BCI is no observable movement of muscles, the definition is not met.<\/p>\n<p>This mismatch showcases the paradigm shift of BCI command: for as long as criminal law has existed, any criminal conduct has always entailed an externally visible bodily act. Where previously, a soldier needed to pull a trigger to shoot a person, BCIs would allow merely thinking about pulling a trigger, and yet shoot a person. As he induces the same effects in the physical world, it would be seriously perplexing if solely by switching to an active BCI the soldier was exonerated from criminal liability. The decisive factor cannot be whether the operator moves his body, but whether he creates a physical <a href=\"https:\/\/link.springer.com\/article\/10.1007\/s12152-019-09416-5\" rel=\"nofollow noopener\" target=\"_blank\">impact<\/a> on the outside world. A consciously produced mental command thus constitutes an act.<\/p>\n<p>But what precisely is the actus reus? Let us assume the BCI user has launched the connected weapon and thereby engaged in the war crime of <a href=\"https:\/\/www.icc-cpi.int\/sites\/default\/files\/2024-05\/Rome-Statute-eng.pdf\" rel=\"nofollow noopener\" target=\"_blank\">directing<\/a> an attack against the civilian population (Rome Statute, art. 8(2)(b)(i)). Has he directed an attack by <a href=\"https:\/\/www.elgaronline.com\/monochap\/book\/9781035394456\/chapter13.xml\" rel=\"nofollow noopener\" target=\"_blank\">activating<\/a> the BCI? By contemplating the attack? By <a href=\"https:\/\/link.springer.com\/article\/10.1007\/s12152-019-09416-5\" rel=\"nofollow noopener\" target=\"_blank\">generating<\/a> the neural signals the computer recognizes? By <a href=\"https:\/\/link.springer.com\/article\/10.1007\/s12152-019-09416-5\" rel=\"nofollow noopener\" target=\"_blank\">evoking<\/a> the electrical activity within the computer to launch the weapon?<\/p>\n<p>Ultimately, this question revolves around the necessary causal nexus between conduct and effect. The doctrinally sound answer points to the act of eliciting the neural signals the computer transmits. In the context of the active BCI, this is the <a href=\"https:\/\/link.springer.com\/article\/10.1007\/s12152-019-09416-5\" rel=\"nofollow noopener\" target=\"_blank\">temporally closest<\/a>, voluntary action in the causal chain, which is necessary to incite the attack.<\/p>\n<p>It remains to be evaluated if the same considerations apply to the reactive BCI. The soldier equally generates neural signals, albeit not consciously. Determinative, therefore, is whether the actus reus must be a voluntary act (based on conscious volition). So far, this question has been mostly academic and will need to be addressed once reactive BCIs for weapon control manifest. If one were to request a <a href=\"https:\/\/academic.oup.com\/oxford-law-pro\/book\/57435\/chapter\/473927308\" rel=\"nofollow noopener\" target=\"_blank\">willed action<\/a>, the last voluntary act with a causal connection to the attack would be the operator activating the BCI in the first place.<\/p>\n<p>This controversy is highly relevant as the perpetrator must entertain the mental element at the time of the actus reus.<\/p>\n<p>Mens Rea <\/p>\n<p>The mens rea requires committing the actus reus with intent, both regarding the conduct and, where applicable, regarding a consequence (see Rome Statute, \u00a0art. 30(1)-(2)).<\/p>\n<p>Sticking to the war crime of directing attacks against the civilian population, the <a href=\"https:\/\/www.icc-cpi.int\/sites\/default\/files\/Publications\/Elements-of-Crimes.pdf\" rel=\"nofollow noopener\" target=\"_blank\">elements of crime<\/a> include the consequence that \u201c[t]he perpetrator intended the civilian population \u2026 to be the object of the attack.\u201d Ergo, the mens rea exists when the operator launches the attack, while meaning to attack civilians or while being aware that civilians would be struck in the ordinary course of events. Presuming that no bi-directional information inflow or interference has affected the soldier\u2019s volition, the active BCI does not preclude the mens rea.<\/p>\n<p>If one, for the reactive BCI, identifies the subconsciously evoked neural signal as the actus reus, intent cannot be established in the same manner. There can be no intent without conscious thought. Conversely, if the actus reus was the activation of the BCI, the soldier may in theory\u2014but likely not in practice\u2014satisfy the mens rea. The threshold for oblique intent (the operator being aware that civilians would be attacked) is \u201c<a href=\"https:\/\/www.icc-cpi.int\/sites\/default\/files\/CourtRecords\/CR2009_04528.PDF\" rel=\"nofollow noopener\" target=\"_blank\">virtual certainty<\/a>.\u201d Hence, the soldier would need to be aware, with virtual certainty, of subconsciously selecting civilian targets. Possibly, a superior commander having <a href=\"https:\/\/www.icc-cpi.int\/sites\/default\/files\/CourtRecords\/CR2009_04528.PDF\" rel=\"nofollow noopener\" target=\"_blank\">failed<\/a> to intervene when the operator was expected to attack civilians could instead be held liable in this instance (Rome Statute, art. 28).<\/p>\n<p>Nonetheless, there might be an accountability gap through the defense of \u201cmy brain did it, not me.\u201d One may or may not consider this an acceptable outcome.<\/p>\n<p>Final Thoughts<\/p>\n<p>In the end, said determinations are not as much a question of lex lata as of lex ferenda. This post attempts no more than to reveal the primary shifts and challenges once BCI-controlled weapon systems emerge. The remedies for such issues might differ substantially based on the ensuing technological specificities. In the concluding observations of his book, Bill Boothby contemplated \u201cwhether future law will need to <a href=\"https:\/\/www.elgaronline.com\/monochap\/book\/9781035394456\/chapter14.xml\" rel=\"nofollow noopener\" target=\"_blank\">distinguish<\/a> between thoughts.\u201d<\/p>\n<p>I suggest that future law will need to distinguish between brain signals. Without question, tackling such issues of neurowarfare compels a serious interdisciplinary effort in understanding how these technologies work and addressing the complications they bear.<\/p>\n<p style=\"text-align: center;\">***<\/p>\n<p><a href=\"https:\/\/lieber.westpoint.edu\/about\/team\/profile\/?smid=40921\" rel=\"nofollow noopener\" target=\"_blank\">Denise Koecke<\/a> holds a degree in International Relations from TU Dresden and is currently a law student at European University Viadrina Frankfurt (Oder).<\/p>\n<p>The views expressed are those of the author, and do not necessarily reflect the official position of the United States Military Academy, Department of the Army, or Department of Defense.\u00a0<\/p>\n<p>Articles of War\u00a0is a forum for professionals to share opinions and cultivate ideas.\u00a0Articles of War\u00a0does not screen articles to fit a\u00a0particular editorial\u00a0agenda, nor endorse or advocate material that is published. Authorship does not indicate affiliation with\u00a0Articles of War, the Lieber Institute, or the United States Military Academy West Point.<\/p>\n<p>\u00a0<\/p>\n<p>\u00a0<\/p>\n<p>\u00a0<\/p>\n<p>\u00a0<\/p>\n<p>\u00a0<\/p>\n<p>\u00a0<\/p>\n<p>\u00a0<\/p>\n<p>\u00a0<\/p>\n<p>Photo credit: Getty Images via Unsplash+<\/p>\n","protected":false},"excerpt":{"rendered":"Editors\u2019 note: This post is an entry in our series on Air Commodore (ret.) William H. Boothby\u2019s book,&hellip;\n","protected":false},"author":2,"featured_media":87603,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[11],"tags":[42852,3956,42853,42854,42855,42856,1469,1785,4120,42857,2653,42858,42859,26615,42860,6924,42861],"class_list":["post-87602","post","type-post","status-publish","format-standard","has-post-thumbnail","category-military-defense","tag-actus-reus","tag-ap-i","tag-bci","tag-brain-technology","tag-denise-koecke","tag-human-control","tag-icl","tag-ihl","tag-international-criminal-law","tag-mens-rea","tag-military-weapons","tag-neural-signal","tag-rome-statute","tag-targeting","tag-targeting-law","tag-weapon","tag-weapon-control"],"share_on_mastodon":{"url":"https:\/\/pubeurope.com\/@conflicts\/117633527662049258","error":""},"_links":{"self":[{"href":"https:\/\/www.europesays.com\/conflicts\/wp-json\/wp\/v2\/posts\/87602","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.europesays.com\/conflicts\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.europesays.com\/conflicts\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/conflicts\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/conflicts\/wp-json\/wp\/v2\/comments?post=87602"}],"version-history":[{"count":0,"href":"https:\/\/www.europesays.com\/conflicts\/wp-json\/wp\/v2\/posts\/87602\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/conflicts\/wp-json\/wp\/v2\/media\/87603"}],"wp:attachment":[{"href":"https:\/\/www.europesays.com\/conflicts\/wp-json\/wp\/v2\/media?parent=87602"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.europesays.com\/conflicts\/wp-json\/wp\/v2\/categories?post=87602"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.europesays.com\/conflicts\/wp-json\/wp\/v2\/tags?post=87602"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}