Scenario 14: Ransomware campaign

CC-0 Pixabay

Municipal governments and health care providers in one State fall victim to a ransomware campaign launched by a non-State group in a second State. The ransomware campaign disables municipal and health care services in the first State. The scenario explores how the ransomware campaign may be classified under international law. It first considers whether the campaign is a breach of an international obligation attributable to a State. It then discusses the possible legal responses available to the victim State.

Scenario

Keywords

Attribution, cyber operation, due diligence, prohibition of intervention, ransomware, retorsion, sovereignty, State responsibility, use of force

Facts

[F1] A previously unknown strain of ransomware is directed toward several municipal governments and a variety of health care services providers in State A through the use of phishing emails. Upon opening the emails by government and health care services employees, computer systems are affected. In a major metropolitan city in State A, the local court is forced offline because the ransomware has encrypted its computer systems and the police are forced to revert to using pen and paper to issue traffic citations. Morever, police are unable to effectuate warrants and ongoing investigations into crimes must be postponed. Thousands of computers at the State A Department of Transportation stop working. Processing of applications for drivers’ licenses and permit renewals is halted. City authorities refuse to pay ransom to the attackers and are forced to spend considerable sums to repair and restore the affected computer systems.

[F2] The same ransomware infects hospital systems in a separate city in State A. Doctors are unable to access patient data stored digitally. Staff resort to using paper charts, transmitting messages in person and being able to perform only basic treatment without access to X-rays or ultrasound scans. Health records system of a major company incorporated in State A is also infected, leaving thousands of patient medical files inaccessible. The inaccessibility of patient data coupled with the disruption to the hospital computer systems results in the inability of the medical staff to perform critical surgeries. Patients are admitted to the emergency rooms when absolutely necessary, but cannot be operated on in a timely manner, resulting in several otherwise preventable injuries, but fortunately no loss of life. Lesser harm is caused to patients who cannot be given necessary medication because their medical records are inaccessible. A significant economic loss is caused by the need to reroute patients to other hospitals.

[F3] After several weeks, the ransomware attacks stop.

[F4] Authorities in State A determine that the ransomware was created by a group of hackers in State B. The hackers’ relationship to State B is not clear. However, the methodology utilized by the hackers bears a striking similarity to a previous cyber operation attributed to State B. Moreover, State B, while formally denying any involvement in the incidents, praises the actions of the hackers as a just and foreseeable reaction to what State B characterizes as State A’s foreign policy misdeeds. State A and State B have strained relations.

[F5] State A indicts the hackers, but State B does not cooperate in extraditing the hackers to State A for prosecution under criminal laws of State A for several reasons. Firstly, State B is prohibited by its constitution from extraditing its citizens for criminal prosecution in other States. Secondly, relations between State A and State B are such that, even in the absence of the foregoing reasons, State B would be disinclined to co-operate with State A. Finally, State media in State B has lauded the actions of the hackers as a just response to State A’s purported misdeeds.

Examples

• Sony Pictures Entertainment attack (2014)
• WannaCry (2017)
• NotPetya (2017)
• Texas Municipality ransomware attack (2019)
• SamSam ransomware attack (2018)
• Springhill Medical Center ransomware attack (2019)
• Colonial Pipeline ransomware attack (2021)
• Kaseya VSA ransomware attack (2021)

Legal analysis

For a general overview of the structure of analysis in this section, see Note on the structure of articles.

[L1] The analysis in this scenario focuses on the legal qualification of the ransomware attacks from the perspective of international law. In particular, it examines whether the relevant conduct is attributable to State B and whether it amounts to a breach of an international obligation owed by State B to State A. It then discusses the possible legal responses available to the State A.

Attribution

State organs and exercise of governmental authority

State organs and persons and entities in exercise of governmental authority

The following types of conduct of State organs and persons and entities in exercise of governmental authority are attributable to a State: The conduct of any of the organs of that State, "whether the organ exercises legislative, executive, judicial or any other functions, whatever position it holds in the organization of the State, and whatever its character as an organ of the central Government or of a territorial unit of the State"; The conduct of "a person or entity which is not an organ of the State […] but which is empowered by the law of that State to exercise elements of the governmental authority, […] provided the person or entity is acting in that capacity in the particular instance"; The conduct of an organ of another State placed at the disposal of the State in question, if "the organ is acting in the exercise of elements of the governmental authority" of the latter State. Such conduct is attributable to the State even if the organ, person or entity acting in that capacity "exceeds its authority or contravenes instructions" (acts ultra vires).

[L2] It is not clear that the hackers are a State organ of State B, nor that they are exercising governmental authority on behalf of State B.

Non-State actors

Non-State actors

Activities of non-State actors (groups and individuals) are generally not attributable to States. However, such conduct can be attributable to a State in particular if the actor is: "in fact acting on the instructions of, or under the direction or control of, that State in carrying out the conduct"; Each of the three criteria entails a form of subordination between the non-State actor and the potentially responsible State. Regarding the criterion of control, there is a debate on the degree of control required for the attribution of the conduct to the State, as different tests have been developed. On the one hand, the ICJ has affirmed that the exercise of “effective control” is necessary, which entails that the State is able to control the beginning of the relevant operations, the way they are carried out, and their end. This position has been expressly followed by some States in the realm of cyber operations, including Brazil, the Netherlands and Norway. On the other hand, a less restrictive approach has been developed by the ICTY, and followed by the ICRC, under the “overall control” test, which requires the State in question (i) to provide the non-State entity with financial and training assistance, military equipment and/or operational support, and (ii) to participate in the organization, co-ordination or planning of operations of the entity in question. Nevertheless, the proponents of this test limit it to organized groups, meaning that the effective control test remains applicable for the conduct of private individuals, or unorganized groups. "in fact exercising elements of the governmental authority in the absence or default of the official authorities and in circumstances such as to call for the exercise of those elements of authority"; "an insurrectional movement which becomes the new Government of a State"; or "a movement, insurrectional or other, which succeeds in establishing a new State in part of the territory of a pre-existing State or in a territory under its administration". Additionally, the conduct of a non-State actor is attributable to a State "if and to the extent that the State acknowledges and adopts the conduct in question as its own".

[L3] As non-State actors, the actions of the hackers may be attributable to State B if they were acting under the instructions or effective control of State B or if State B acknowledges and adopts the actions as its own. First, there is no evidence to suggest that the hackers were operating under the instructions or effective control of State B. Likewise, there is insufficient evidence to indicate that State B adequately acknowledged and adopted the behaviour of the hackers as its own. Although State B, including its State media, praised the actions of the hackers as a just and foreseeable reaction to what State B characterizes as State A’s foreign policy, this is unlikely to satisfy the acknowledgment and adoption standard which is applied narrowly and requires more State involvement that mere endorsement or expression of approval. Moreover, a minority view holds that the acknowledgement and adoption standard only applies prospectively, and, thus, under this view, State B’s praise of the actions of the hackers after the fact would not suffice to meet the standard.

Breach of an international obligation

[L4] This section considers whether the ransomware attacks are a breach of an international obligation—specifically, the prohibition on the use of force, the prohibition on intervention, the obligation to respect the sovereignty of other States, and the duty of due diligence.

Use of force

Use of force

Article 2(4) of the UN Charter prescribes States to “refrain in their international relations from the threat or use of force against the territorial integrity or political independence of any state, or in any other manner inconsistent with the purposes of the United Nations”. This prohibition is reflective of customary international law and it is frequently described as a peremptory norm of international law. This rule applies between States; therefore the conduct needs to be attributable to a State and against another State ‘in their international relations’, thus excluding non-State actors unless their conduct is attributable to a State. As stated by the International Court of Justice, the prohibition applies to any use of force, regardless of the means employed. However, the notion of “force” in this context is limited to armed force, and to operations whose scale and effects are comparable to the use of armed force. As stressed by several States, each situation has to be analysed on a case-by-case basis. Undoubtedly, one of the purposes of the prohibition of force under international law is to safeguard the national security of the potentially affected States. However, many forms of outside interference including various forms of political and economic coercion may affect the national security of the victim State. And yet, the drafters of the UN Charter had expressly rejected the proposal to extend the prohibition of force beyond the strict confines of military (or armed) force. This is reflected also in the preamble, which explicitly stipulates that the drafters sought “to ensure, by the acceptance of principles and the institution of methods, that armed force shall not be used, save in the common interest”. In principle, it could be argued that the notion of “force”, like other generic terms in treaties of unlimited duration, should be presumed to have an evolving meaning. Regarding its application to cyber operations, an “effects-based approach” has been mostly followed. In this sense, there is emerging consensus that “a cyber attack that causes or is reasonably likely to cause physical damage to property, loss of life or injury to persons would fall under the prohibition contained in Article 2(4) of the UN Charter”, including both direct and indirect consequences. At present, there is a debate as to whether cyber operations with no physical effects may amount to a prohibited use of force. It has been argued that disruptive cyber operations of this kind fall under the scope of Article 2(4) if the resulting disruption is “significant enough to affect state security”. As of 2022, there is limited State practice supporting the claim that the meaning of “force” has evolved to include non-destructive cyber operations against critical national infrastructure and no victim State of an operation of this kind has suggested that the operation would have amounted to a use of force. However, States have begun addressing this question. In particular, France, the Netherlands and Norway allow for the possibility of cyber operations, which do not produce physical effects, to qualify as uses of force, if certain criteria are met. These qualitative and quantitative non-exhaustive criteria include the seriousness and reach of a given cyber operation’s consequences and its military nature, as well as “the circumstances prevailing at the time of the operation, such as the origin of the operation and the nature of the instigator (military or not), the extent of intrusion, the actual or intended effects of the operation or the nature of the intended target”. Several of these criteria are also reflected in the Tallinn Manual 2.0. Other States, such as Italy, did not rule out the possibility of considering operations causing the interruption of essential services without physical damage within the scope of the prohibition of the use of force. A use of force is unlawful under international law, unless it is authorized by the UN Security Council under Chapter VII of the UN Charter, conducted in the exercise of the inherent right to self-defence, or consented to by the territorial State. Even if an operation does not meet the threshold of the use of force, it may still be considered a violation of other rules of international law. In this regard, the prohibition of intervention, the obligation to respect the sovereignty of other States, and the possible obligation to refrain from launching cyber operations against other States’ critical infrastructure are all of potential relevance. Publicly available national positions that address this issue include: (2024), (2020), (2021), (2022), (2023), (2023), (2020), (2019), (2021), (2023), (2020), (2021), (2021), (2019), (2021), (2023), (2022), (2021), (2022), (2021), (2012), (2020), (2021).

[L5] It is unlikely that the ransomware attacks amount to a use of force. A consideration of eight factors proposed by the Tallinn Manual — severity, immediacy, directness, invasiveness, measurability of effects, military character, State involvement, and presumptive legality — shows that to be so. However, the eight factors do not constitute an official framework that must be used when assessing the threshold of use of force, but are rather non-exhaustive indicators, similar to those that States are likely to consider.

[L6] Crucially, the ransomware attacks’ overall severity was significant. While the disruption in municipal functions caused only inconvenience for citizens and municipal agencies, the harm that resulted from the disruption to medical services was significant. Several patients admitted to hospitals in State A for emergency procedures were unable to be operated on in a timely manner, resulting in several instances in otherwise preventable injury. Other patients could not receive their medication due to the inaccessibility of their medical records.

[L7] The consequences of the ransomware attacks did not follow immediately from the cyber activities. In most cases, the penetration of the affected systems occurred weeks before the ransom notice was directed to the victim, and monetary costs incurred by the victims to recover data and restore their systems followed weeks or months thereafter. States are more likely to view a cyber operation that causes immediate consequences as a use of force, as distinguishable from consequences that are delayed or manifest slowly over time. Likewise, the less opportunity a State has to mitigate or forestall the effects of a cyber operation, the more likely it is that the operation will be considered as a use of force. State A, in fact, had the choice of making a ransom payment and thus preventing the consequences of the attacks.

[L8] The effects of the ransomware attacks on State A were not directly connected to the underlying cyber activity. Directness depends on the degree of attenuation between the initial act and its consequences. The initial act in this case was the sending of phishing emails. When those emails were opened, the ransomware infected the recipient’s computers. Computer files were then encrypted and rendered inaccessible. Because the victims failed to make ransom payments, the files were not readily decrypted. Inability to access the files ultimately and indirectly resulted in harm to the victim organizations. While the attacks did have indirect consequences, in the form of the costs incurred to restore backed-up data and to implement improved security, the degree of attenuation between the initial act and its consequences means that directness of the attacks’ cause and effects may not be viewed as comparable to the direct harm caused to people or objects by a kinetic attack.

[L9] The hackers did indeed invasively probe the networks of municipal governments and healthcare providers; however, these were not top-secret networks that were necessarily intended to have the highest level of security. And the networks that the hackers did access were not amongst the most secure maintained by the victims: for instance, emergency response networks were untouched. The effects of the ransomware attacks cannot be calculated with certainty, even if a numerical sum can be affixed to the remediation costs.

[L10] There is no indication that the attacks had a military character. No link has been established between the hackers and the military of State B or of any other State. Nor were the military forces of State A the target of the ransomware campaign.

[L11] Likewise, no State is publicly alleged to have been involved, either directly or indirectly, in the campaign. In this regard, the lack of a clear attribution of the cyber operation to a State would also preclude the characterization of the operation as a use of force, which applies only between States.

[L12] Finally, the reconnaissance and network probing activities of the hackers were qualitatively similar to espionage activities, which are not per se regulated under international law and thus not presumptively judged to be uses of force.

[L13] Considering each one of the foregoing factors, severity is the only one that could feasibly cause the ransomware attacks to meet the criteria of a use of force. However, severity is subject to a de minimis rule. It is likely that the severity of the ransomware attacks was not sufficient to meet that de minimis. While the otherwise preventable injuries to persons are certainly more than mere inconveniences, it is unclear whether the limited number of injuries, in light of the ransomware attacks’ failure to meet any of the other criteria, renders the attacks so severe as to constitute a use of force. However, scholars and States may reach different conclusions in their assessment.

Prohibition of intervention

Prohibition of intervention

The obligation of non-intervention, a norm of customary international law, prohibits States from intervening coercively in the internal or external affairs of other States. Prohibited intervention was authoritatively defined by the International Court of Justice in the judgment on the merits in the 1986 Nicaragua v United States case: A prohibited intervention must […] be one bearing on matters in which each State is permitted, by the principle of State sovereignty, to decide freely. One of these is the choice of a political, economic, social and cultural system, and the formulation of foreign policy. Intervention is wrongful when it uses methods of coercion in regard to such choices, which must remain free ones. In order for an act, including a cyber operation, to qualify as a prohibited intervention, it must fulfil the following conditions: The act must bear on those matters in which States may decide freely. The spectrum of such issues is particularly broad and it includes both internal affairs (such as the “choice of a political, economic, social, and cultural system” or the conduct of national elections), and external affairs (“formulation of foreign policy”; or “recognition of states and membership of international organisations”)—the so-called domaine réservé of States. The content of the domaine réservé is determined by the scope and nature of the State's international legal obligations. The act must be coercive in nature. There is no generally accepted definition of “coercion” in international law. In this respect, two main approaches have emerged in the cyber context: Under the first approach, an act is coercive if it is specifically designed to compel the victim State to change its behaviour with respect to a matter within its domaine reservé. Under this approach, the “key is that the coercive act must have the potential for compelling the target State to engage in an action that it would otherwise not take (or refrain from taking an action it would otherwise take)”. Under the second approach giving meaning to “coercion”, it is sufficient for an act to effectively deprive the target State of its ability to control or govern matters within its domaine reservé. This latter approach distinguishes itself from the former by accepting that mere deprivation of the target State’s control over a protected matter, without actually or potentially compelling that State to change its behaviour, may constitute intervention. Under both approaches, however, merely influencing the target State by persuasion or propaganda or causing a nuisance without any particular goal is insufficient to qualify as coercion. The element of coercion also entails the requirement of intent. While coercion is evident in the case of an intervention involving the use of force, ‘either in the direct form of military action, or in the indirect form of support for subversive or terrorist armed activities within another State’, as affirmed by the ICJ, it is less clear with respect to non-forcible forms of interference. Some States support the approach that intervention may take various forms, such as economic and political coercion. One example that has been reiterated in several States’ positions, including Australia, Brazil, Canada, Germany, Israel, New Zealand, Norway, Singapore, the United Kingdom and the United States, is the case of cyber operations by a State interfering with another state’s ability to hold an election or manipulating the election results. Many States have affirmed that the assessment has to be done on a case-by-case basis. Both potential and actual effects are considered to be relevant when assessing the coercion element. Finally, there has to be a causal nexus between the coercive act and the effect on the internal or external affairs of the target State. The prohibition of intervention applies between States, and thus it is not applicable to the activities of non-State groups, unless their conduct can be attributed to a State under the rules on attribution under international law. Publicly available national positions that address this issue include: (2024), (2020), (2021), (2022), (2023), (2021), (2023), (2021), (2019), (2021), (2020), (2023), (2020), (2021), (2021), (2019), (2020), (2021), (2023), (2022), (2021), (2021), (2022), (2021), (2018), (2021), (2022), (2016), (2020), (2021).

[L14] There is no suggestion that the ransomware attacks in any way involved the external affairs of State A, but certain effects of the ransomware did implicate the domaine réservé of State A. For example, the conduct of traffic police, the operations of the Department of Transportation, or the provision of essential medical care, are certainly fields of activity not committed to international law and possibly form part of the domaine réservé of State A.

[L15] It is less likely that the attacks were coercive efforts designed to influence outcomes in those fields of activity. While the hackers may have manipulated hospitals in State A and the municipal government of a city in State A into making a choice between paying a ransom or spending considerably more to remedy the effects, that choice was not coercive in the sense that it was designed to compel the State A to adopt a particular policy, or to deprive it of its ability to control its affairs, with regard to traffic, policing, hospitals, or municipal policy. Instead, the coercion was intended to compel the payment of ransom.

[L16] Accordingly, the ransomware attacks would likely not be considered a coercive intervention in the domaine réservé of State A.

Obligation to respect the sovereignty of other States

Sovereignty

Sovereignty is a core principle of international law. According to a widely accepted definition of the term in the 1928 Island of Palmas arbitral award, [s]overeignty in the relations between States signifies independence. Independence in regard to a portion of the globe is the right to exercise therein, to the exclusion of any other State, the functions of a State. Multiple declarations by the UN, the African Union, the European Union, NATO, OSCE, and individual States have confirmed that international law applies in cyberspace. Accordingly, so too does the principle of sovereignty. However, there is some debate as to whether this principle operates as a standalone rule of international law, the breach of which gives rise to state responsibility. For the proponents of this view, the prohibition on violating the sovereignty of other States is a substantive primary rule of international law, the breach of which is an internationally wrongful act. This view was unanimously accepted by the experts who prepared the Tallinn Manual 2.0. It has also been adopted by several States including Austria, Brazil, Canada, the Czech Republic, Estonia, Finland, France, Germany, Iran, Italy, Japan, the Netherlands, New Zealand, Norway, Romania and Sweden. By contrast, the opposing view is that sovereignty is a principle of international law that may guide State interactions, but it does not amount to a standalone primary rule. This view has been adopted by one State, the United Kingdom, and has been partially endorsed by the U.S. Department of Defense General Counsel. By this approach, cyber operations cannot violate sovereignty as a rule of international law, although they may constitute prohibited intervention, use of force, or other internationally wrongful acts. The remainder of this section proceeds on the basis of the former “sovereignty-as-rule” approach. Those espousing the latter “sovereignty-as-principle” approach should refer to other relevant sections of the legal analysis (such as that on the prohibition of intervention or use of force). It is understood that sovereignty has both an internal and an external component. In the cyber context, the “internal” facet of sovereignty entails that “[a] State enjoys sovereign authority with regard to the cyber infrastructure, persons, and cyber activities located within its territory, subject to its international legal obligations.” This encompasses both private and public infrastructure. The external component entails that States are “free to conduct cyber activities in [their] international relations”, subject to their international law obligations. As a general rule, each State must respect the sovereignty of other States.However, within the cyber realm – and particularly regarding remote cyber operations – there is still no agreement on the criteria and the required threshold to qualify an operation as a sovereignty violation. It is clear that a cyber operation with severe destructive effects, comparable to a “non-cyber” armed attack or a use of force against a State, constitutes a violation of its sovereignty; however, with more subtle cyber operations, the question is far from settled. Accordingly, the assessment needs to be done on a case-by-case basis. The following modalities, highlighted in the Tallinn Manual 2.0, represent different ways of determining what a “sovereignty violation” might mean in the context of cyber operations: A State organ conducting cyber operations against a target State or entities or persons located there while physically present in the target State's territory violates the target State's sovereignty. This was agreed by all Experts drafting the Manual; however, “a few” of the Experts thought that the extensive State practice carved out an exception for espionage operations. Causation of physical damage or injury by remote means; again, “a few” Experts took the position that this is a relevant but not a determinative factor by itself. Causation of a loss of functionality of cyber infrastructure: although the Tallinn Manual 2.0 experts agreed that a loss of functionality constituted “damage” and thus a breach of sovereignty, no consensus could be achieved as on the precise threshold for a loss of functionality (the necessity of reinstallation of the operating system or other software was proposed but not universally accepted); Below this threshold, there was no agreement among the Experts whether operations that do not cause physical consequences or a loss of functionality qualify as a violation of sovereignty. Interference with data or services that are necessary for the exercise of "inherently governmental functions"; although the Experts could not conclusively define the term "inherently governmental functions", they agreed that, for example, the conduct of elections would so qualify. Usurpation of "inherently governmental functions", such as exercise of law enforcement functions in another State’s territory without justification. The Tallinn Manual’s view of what constitutes a violation of sovereignty has been expressly endorsed by several States including Canada, Germany and the Netherlands; and followed to some extent by other States, such as the Czech Republic, Norway, Sweden and Switzerland. An alternative test has been proposed by France, which argues that a breach of sovereignty occurs already when there is “any unauthorised penetration by a State of [the victim State’s] systems”;similarly, Iran has argued that “unlawful intrusion to the (public or private) cyber structures” abroad may qualify as a breach of sovereignty. Attributing the relevant cyber operation to a State different from the target State is a necessary prerequisite for qualifying the cyber operation as a violation of the target State's sovereignty. Whether non-State actors can violate territorial sovereignty on their own is a matter of disagreement. Publicly available national positions that address this issue include: (2024), (2020), (2021), (2022), (2021), (2023), (2020), (2023), (2019), (2021), (2020), (2019), (2021), (2020), (2023), (2020), (2021), (2021), (2021), (2019), (2020), (2021), (2023), (2022), (2021), (2021), (2022), (2021), (2018), (2021), (2022), (2012), (2016), (2020), (2021).

[L17] The attacks caused injury to persons (option 2 above), and they resulted in severe losses of functionality (option 3). Everything from normal employee workstations at the Department of Transportation to sophisticated surgical systems and diagnostic equipment at the hospital were unable to function. Medical services were disrupted. Municipal offices were forced offline for weeks. The loss of functionality required spending considerable sums of money on IT support and repair and reinstallation of software to remediate.

[L18] Moreover, the ransomware attacks also interfered with the performance of inherently governmental functions (option 4 above). The court and police operations of State A’s cities are inherently governmental functions, which although not usurped were certainly subject to interference.

[L19] Thus, provided the ransomware attacks were attributable to State B (on which, see above), they would likely amount to a violation of the State A’s sovereignty.

Due diligence obligation

Due diligence

According to the traditional formulation by the ICJ in the Corfu Channel case, every State is under an “obligation not to allow knowingly its territory to be used for acts contrary to the rights of other States”. In the cyber context, the UN General Assembly urged States already in 2000 to “ensure that their laws and practice eliminate safe havens for those who criminally misuse information technologies”. It is the matter of some controversy whether the principle of due diligence reflects a binding obligation applicable to cyber operations. Some States have framed it within their national positions as one of the "voluntary, non-binding norms of responsible State behaviour" in cyberspace, including Israel, New Zealand, the United Kingdom and Canada. It has also been proposed that in the cyber context, it is preferable to construe due diligence as a standard of attribution rather than as a standalone primary rule of international law. Nevertheless, the present analysis proceeds on the basis that as a matter of lex lata, due diligence constitutes a general international obligation for every State not to knowingly allow its territory to be used for internationally wrongful acts using cyber means. This view has also been endorsed by a growing number of States, including Australia, Czech Republic, Estonia, Finland, France, Germany, Italy, Japan, the Netherlands, Norway, Switzerland, and Sweden. Due diligence does not entail a duty of prevention, but rather an obligation of conduct. A State breaches its due diligence obligation in the presence of the following cumulative elements: The existence of acts (by a non-State actor or a third State) contrary to the rights of a victim State, which are conducted from or through the territory of the potentially responsible State (or from or through the territory or cyber infrastructure under its control), which would have been unlawful if conducted by the potentially responsible State, which have serious adverse consequences for the victim State, with respect to which the potentially responsible State has actual or constructive knowledge, and upon which the potentially responsible State can act, but fails to take all feasible measures. Publicly available national positions that address this issue include: (2024), (2020), (2022), (2021), (2023), (2020), (2023), (2019), (2021), (2019), (2021), (2023), (2020), (2021), (2021), (2019), (2020), (2021), (2022), (2021), (2021), (2022), (2021), (2021), (2021).

[L20] Breaches of the due diligence obligation do not require that the act in question be attributable to a State. Instead, the duty of due diligence assumes the role of three parties: the target State toward which the cyber operation is directed; the territorial State; and a third-party author of the cyber operation. The third party may be another State, a non-State group, or a private person. The threshold is whether the underlying act, if committed by the territorial State, would have violated a rule of international law (i.e., would have been considered an internationally wrongful act) (condition 3). While, in general, it is States, not individuals or private entities, that are able to violate international law, cyber operations carried out by individuals or private entities that nevertheless result in serious adverse consequences fall within a State’s due diligence obligation.

[L21] Thus, if State B has knowledge of the hackers’ operations, those operations affect the rights of and cause serious adverse consequences to State A, and State A intimates that State B take action to stop the interference in its internal affairs, State B has a duty to take feasible action to stop the ransomware attacks. While the harm caused by a cyber activity must be serious, the due diligence principle does not require that there be physical damage to objects or injuries to persons.

[L22] In the present case, the ransomware attacks affected State A’s sovereign right to perform inherently governmental functions—operating courts and police departments (condition 1). Moreover, the attacks caused serious adverse consequences for the victim State. The loss of functionality in hospital equipment caused injury to patients, and the encryption of health records led to improper medical care resulting in injury to patients (condition 4).

[L23] Regarding the required knowledge, a State is in breach even if it is unaware of cyber activity conducted from its territory but “objectively should have known that its territory was being used” (condition 5). State A believes that State B knew that the hackers were operating from its territory (conditions 2 and 5). State B, on the other hand, argues that it had no such knowledge. It is not reasonable to assert that State B objectively should have known that the hackers were utilizing its cyber infrastructure to launch attacks against State A because the ransomware employed was complex and previously unknown. State B cannot be expected to prevent the use of malware of which is it entirely unaware. Given that the knowledge requirement was not clearly met in this case, State B’s lack of response did not amount itself to a violation of its duty of due diligence (condition 6).

Permissible responses by State A

[L24] Having established that the ransomware attacks do not meet the criteria of an internationally wrongful act, this section examines the options available for State B to take in response. Cyber operations may, in general, be met with four responses under international law: countermeasures, the plea of necessity, self-defense, and retorsion. For the reasons explained below, only retorsion is available to State A.

Countermeasures

Countermeasures

Countermeasures are “measures that would otherwise be contrary to the international obligations of an injured State vis-à-vis the responsible State, if they were not taken by the former in response to an internationally wrongful act by the latter in order to procure cessation and reparation”. Several States, including Australia, Austria, Canada, Estonia, France, Germany, Italy, Japan, the Netherlands, New Zealand, Norway, Singapore, Sweden, the United Kingdom, and the United States, have expressly confirmed the applicability of the law of countermeasures to cyber operations. Others, including Brazil, China, and Cuba, have expressed caution in this regard. Countermeasures should be distinguished from retorsions, which are unfriendly but lawful acts by the aggrieved party against the wrongdoer. As a matter of general international law, an injured State may only take countermeasures against the responsible State if the following conditions are met: The existence of a prior internationally wrongful act of the responsible State against the injured State. If that act consists of a cyber operation, this means that the operation must have been amounted to a breach of the obligations of the responsible State that is attributable to that State; The injured State has called upon the responsible State to fulfil its obligations arising from its internationally wrongful act; and The injured State has notified the responsible State of its decision to take countermeasures, and offered to negotiate with that State, unless it is taking “urgent countermeasures as are necessary to preserve its rights.” Some States, such as Canada, Israel, Norway, the United Kingdom and the United States have advocated for a particular approach to the notification requirement in the cyber realm, in particular to preserve the effectiveness of the measures and/or to avoid exposing sensitive capabilities of the responding State. Additionally, the countermeasures must fulfil the following requirements: Their aim must be to induce the responsible State to comply with the legal consequences of its internationally wrongful act; hence, the aim of countermeasures is restoration, not retribution or retaliation; and the countermeasures can only target the responsible State; They “shall, as far as possible, be taken in such a way as to permit the resumption of performance of the obligations in question”; They shall not affect the obligation to refrain from the threat or use of force, obligations for the protection of fundamental human rights, of a humanitarian character prohibiting reprisals, or other “obligations under peremptory norms of general international law”; obligations under any dispute settlement procedure between the injured and responsible State, and obligations arising from the inviolability of diplomatic or consular agents, premises, archives and documents; and They must be “commensurate with the injury suffered, taking into account the gravity” of the prior unlawful act and of the rights in question (i.e. the “proportionality” requirement). However, proportionality does not require that the adopted measures must be equivalent, reciprocal or even in kind. As clearly expressesd by many States, including Canada, Germany, Italy, Japan, Norway, Sweden, Switzerland, the United Kingdom and the United States, countermeasures against cyber operations can be non-cyber in nature, and cyber countermeasures may be adopted in response to non-cyber wrongful acts. Countermeasures are temporary in nature. In case the original internationally wrongful act has ceased, and the dispute is submitted in good faith to a court or tribunal with the authority to make decisions binding on the parties, countermeasures may not be taken and if already taken, must be suspended, except if the responsible State fails to implement the dispute settlement procedures in good faith. Countermeasures must also be terminated as soon as the responsible State has complied with its (secondary) obligations. There is an ongoing debate as to whether States that have not themselves been directly injured by an unlawful cyber operation may engage in countermeasures in support of the injured State (sometimes referred to as "collective countermeasures"). In particular, Estonia has opined that non-injured States “may apply countermeasures to support the state directly affected by the malicious cyber operation”, a view that has also received some support from New Zealand. This would apply where diplomatic action is insufficient, but no lawful recourse to use of force exists. This interpretation would allow States to offer active assistance to other States, which may not possess sufficient cyber capabilities themselves to counter an ongoing unlawful cyber operation, or otherwise deter the responsible State with other non-cyber countermeasures. However, this view has since been rejected by at least one State (France), while others, such as Canada, do not yet consider there to be “sufficient State practice or opinio juris to conclude that [collective countermeasures] are permitted under international law”. Therefore, it has to be regarded as a call for progressive development of international law, rather than a statement of the current state of international law. Whether a particular measure fulfils these conditions is an objective question, while the burden of proof that the relevant conditions have been fulfilled falls on the injured State. The exact standard of proof required is unsettled in international law and it will depend on the relevant forum. However, relevant international jurisprudence tends to rely on the standard of “clear and convincing evidence”. This standard translates in practice into a duty to “convince the arbiter in question that it is substantially more likely than not that the factual claims that have been made are true”. Importantly, if a State does resort to countermeasures on the basis of an unfounded assessment that a breach has occurred, it may incur in international responsibility for its own wrongful conduct. Publicly available national positions that address this issue include: (2020), (2021), (2022), (2023), (2023), (2019), (2021), (2020), (2019), (2021), (2023), (2020), (2021), (2021), (2019), (2020), (2021), (2022), (2021), (2021), (2021), (2022), (2021), (2018), (2021), (2022), (2016), (2020), (2021).

[L25] Countermeasures are not an available response to the ransomware attacks in the present scenario because the object of countermeasures must be a State, and, as seen above, it was not possible to attribute these ransomware attacks to a State. Moreover, there must be an internationally wrongful act to justify countermeasures. Even if there was one, countermeasures should be limited to ensuring that the unlawful act stops, potentially obtaining assurance and guarantees of non-repetition from the responsible State, and compelling the responsible State to make reparations. Because the ransomware incidents have stopped, countermeasures would have to be limited to compelling the responsible State to guarantee that the incidents do not resume and providing compensation for damages. Countermeasures may not be punitive or have a retaliatory effect.

[L26] State A would be advised to take precautions not to violate an obligation owed to a third State if engaging in countermeasures in response to the ransomware attacks even if they were attributable to State B because if the countermeasures were to violate a legal obligation owed to a third State, State A could itself be in breach of international law. The wrongfulness of such a breach is not precluded by the validity of the countermeasure against the responsible State. Thus, State A could find itself in breach of its international law obligations.

Plea of necessity

Plea of necessity

The plea of necessity allows a State to act in exceptional cases when there is grave and imminent peril to an essential interest of the State and the action in question is the sole means of safeguarding that interest. Even then, the plea of necessity requires that the injured State’s action be balanced with the interests of any States that would be affected and with those of the international community. The injured State’s action must not seriously impair the essential interests of any other State. The plea of necessity is not available to injured States that have substantially contributed to the situation in which they find themselves. However, the plea of necessity can be asserted to take action against non-State actors and can justify actions that violate the rights of non-responsible States, if these conditions are met. In case the situation of necessity is caused by the conduct of non-State actors, attribution of this conduct to a State is not a precondition for acting based on the plea of necessity. Publicly available national positions that address this issue include: (2019), (2021), (2021), (2019), (2021), (2022), (2021).

[L27] A State’s “essential interest” is not clearly defined. It would certainly include healthcare, justice, and policing. Thus, the ransomware attacks on healthcare service providers and police and court systems certainly impaired essential interests of State A. It is unlikely that the temporary interruption in functionality the ransomware caused was sufficient to put those essential interests in grave and imminent peril. It is also unlikely that no other means existed to safeguard those interests. In any case, the ransomware attacks have abated, if temporarily, and the plea of necessity could only be invoked to end the harmful activity.

Self-defence

Self-defence

A State may respond with force to a cyber operation that qualifies as an “armed attack” pursuant to the customary right to self-defence, as codified in Article 51 of the UN Charter. Most commentators consider only grave uses of force – typically, those that kill or injure persons or damage or destroy property – to constitute armed attacks. The United States, however, takes an outlier position, consistently arguing that any illegal use of force gives rise to the use of force in self-defence. In Nicaragua, the ICJ identified “scale and effects” as criteria upon which to judge whether a use of force constitutes an armed attack. In the Court’s view, only “the most grave” uses of force do so. Thus, only cyber operations that seriously injure or kill a number of persons or cause significant damage to, or destruction of, property would undoubtedly constitute armed attacks. Publicly available national positions that address this issue include: (2020), (2021), (2022), (2023), (2021), (2020), (2019), (2021), (2020), (2023), (2020), (2021), (2021), (2019), (2020), (2021), (2022), (2021), (2022), (2021), (2018), (2021), (2012), (2020), (2021).

[L28] The ransomware attacks did not amount to an armed attack. As discussed previously, the attacks were not even a use of force. Based on a consideration of following eight factors — severity, immediacy, directness, invasiveness, measurability of effects, military character, State involvement, and presumptive legality — the ransomware campaign is not at the level of a use of force (see above). Because the ransomware attacks failed to meet the criteria of use of force they could equally not amount to an armed attack triggering the right to self-defense. State A therefore has no right under international law to respond to the attacks with force.

Retorsion

Retorsion

An act of retorsion is “an unfriendly but nevertheless lawful act by the aggrieved party against the wrongdoer”. Such acts may include the prohibition of or limitations upon normal diplomatic relations, the imposition of trade embargoes or the withdrawal of voluntary aid programmes. Cyber-specific retorsions may include sending warnings to cyber operatives belonging to another State, observing the adversary’s cyber activities on one’s own network using tools such as “honeypots”, or slowing down malicious cyber operations conducted by other States. Publicly available national positions that address this issue include: (2023), (2021), (2021), (2019), (2020), (2021), (2022), (2021), (2021), (2022), (2016), (2021).

[L29] Retorsion is the only way for State A to respond to the ransomware campaign without a determination that another State B has breached an international obligation owed to it. State A is likely to view State B’s refusal to extradite or prosecute the attack’s perpetrators as an unfriendly act. State A may signal its displeasure at State B’s unfriendly action through common methods of retorsion such as expelling State B’s diplomats or imposing financial sanctions on State B.

Checklist

• Can the ransomware attacks be attributed?
  • Are the authors of the ransomware attack State organs or non-State organs exercising governmental authority?
  • Are the authors of the ransomware attack non-State actors acting under the instructions or effective control of another State?
• Is the ransomware a use of force?
  • What was the severity of the ransomware attacks?
  • Did the effects of the ransomware attacks manifest immediately?
  • Were the effects of the ransomware attacks directly connected to the underlying cyber activity?
  • Did the ransomware attacks involve invasive penetration of top secret networks?
  • Were the effects of the ransomware attacks measurable?
  • Did the ransomware attacks have a military character?
  • Was a State involved in carrying out or supporting the ransomware attacks?
  • Are the ransomware attacks presumptively legal under international law?
• Is the ransomware a prohibited intervention?
  • Did the ransomware campaign bear on the internal or external affairs of State A?
  • Did the ransomware campaign coerce State A by depriving it of its freedom of choice concerning its internal or external affairs?
• Is the ransomware a violation of the targeted State’s sovereignty?
  • Was the ransomware campaign conducted by a State organ of State B physically present on the territory of State A?
  • Did the ransomware campaign result in physical damage or injury on State A’s territory?
  • Did the ransomware campaign cause a loss of functionality of State A’s computer systems?
  • Did the ransomware campaign interfere with State A’s inherently governmental functions?
• Due diligence
  • If the ransomware attacks were committed by the territorial State, would they have violated a rule of international law?
  • Did the territorial State have knowledge of the attacks?
• What responses are permissible?
  • Is the object of countermeasures a State?
  • Was an essential interest of the victim State in grave and imminent peril?
  • Did the ransomware attacks constitute an armed attack, thus justifying resort to force in self-defense?

Appendixes

See also

• Attribution
• Countermeasures
• Due diligence
• Use of force
• Plea of necessity
• Prohibition of intervention
• Retorsion
• Self-defence
• Sovereignty
• Scenario 03: Cyber operation against the power grid
• Scenario 05: State investigates and responds to cyber operations against private actors in its territory
• Scenario 06: Cyber countermeasures against an enabling State

Bibliography and further reading

• William Banks, State Responsibility and Attribution of Cyber Intrusions After Tallinn 2.0, 95 Tex. L. Rev. 1487 (2017)
• Luke Chircop, ‘A Due Diligence Standard of Attribution in Cyberspace’ (2018) 67 ICLQ.
• James Crawford, State Responsibility: The General Part (CUP 2013).
• Dan Efrony and Yuval Shany, 'A Rule Book on the Shelf? Tallinn Manual 2.0 on Cyberoperations and Subsequent State Practice' (2018) 112 AJIL.
• Robin Geiss and Henning Lahmann, ‘Freedom and Security in Cyberspace: Shifting the Focus Away from Military Responses Towards Non-Forcible Countermeasures and Collective Threat-Prevention’ in Katharina Ziolkowski (ed), Peacetime Regime for State Activities in Cyberspace (NATO CCD COE 2013).
• ET Jensen and S Watts, ‘A Cyber Duty of Due Diligence: Gentle Civilizer or Crude Destabilizer’ (2017) 95 Tex. L. Rev. 1555
• Christian Payne & Lorraine Finlay, Addressing Obstacles to Cyber-Attribution, 49 Geo. Wash. Int’l L. Rev. 535 (2017).
• Marco Roscini, Cyber Operations and the Use of Force in International Law (OUP 2014).
• Michael N. Schmitt, ‘“Below the Threshold” Cyber Operations: The Countermeasures Response Option and International Law’, 54 Va. J. Int’l L. 697 (2014).
• Michael N Schmitt (ed), Tallinn Manual 2.0 on the International Law Applicable to Cyber Operations (CUP 2017).
• US Department of Defense, Office of the General Counsel, Law of War Manual (June 2015)

Contributions

• Scenario by: Kenneth Kraszewski
• Analysis by: Kenneth Kraszewski
• Reviewed by: Anna-Maria Osula, Peter Stockburger, David Wallace

Previous: Scenario 13: Armed conflict

Next: Scenario 15: Cyber deception