Scenario 12: Cyber operations against computer data

In the context of an armed conflict, one belligerent conducts a series of cyber operations against the datasets associated with the other belligerent. These include data used for military purposes, essential civilian datasets, and data serving the enemy’s propaganda. The analysis in this scenario considers the lawfulness of cyber operations designed to corrupt or delete various types of datasets under the law of armed conflict. It particularly focusses on the question whether data qualifies as an “object” for the purposes of the law of armed conflict and whether, as such, it comes within the definition of a military objective.

Keywords
Computer data, distinction, international humanitarian law, military objectives, object, targeting

Facts
State A is involved in a non-international armed conflict against organized armed group G based in its territory. In addition to kinetic hostilities between the two belligerent parties, armed group G conducts a series of cyber operations as part of its military efforts (these are referred to below as incidents 1, 2, and 3):
 * 1) Armed group G conducts a cyber operation against data stored in the computer network at State A’s central military command. The operation results in the deletion or corruption of all data stored in the network, which contained the identity, location, physical condition, staffing, and battle readiness of State A’s warships and military aircraft.
 * 2) Armed group G conducts a cyber operation against data held by State A’s central registry office, a governmental authority maintaining digital records on all State A’s citizens concerning non-military purposes, including census taking, the provision of social benefits, voting, and taxation. The operation results in the deletion of all data held by the office.
 * 3) Armed group G conducts a cyber operation against data stored on the servers of State A’s main press agency. As a result of the operation, all data on the servers of the press agency are deleted and its websites are populated instead with videos and texts calling on the supporters of the regime to resign and defect to the insurgents’ side.

Examples

 * NotPetya (mock ransomware) (2017)

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

The legal analysis first considers whether the definition of military objectives under IHL applies to the situation described in the scenario. It then examines whether computer data qualifies as an “object” under IHL and whether operations against data must therefore be assessed with reference to the elements of that definition. The analysis outlines the two main approaches to that question and notes the extent to which the law is unsettled.

Cyber operations during armed conflicts and the legal definition of military objectives
The central legal issue in the present scenario is whether the deletion or corruption of the specific types of data identified in incidents 1–3 would be lawful under IHL. The present scenario occurs in the context of a NIAC between State A’s government forces and organized armed group G. In accordance with the principle of distinction, only military objectives may be targeted during armed conflict. The definition of military objectives applies in the same manner in an IAC as in a NIAC, for the reasons detailed above. It would thus equally apply to cyber operations executed in the context of the armed conflict between State A and armed group G. Determining the lawfulness of these operations in the present scenario turns predominantly on the question whether data can be considered a military objective.

Qualification of data as a military objective under IHL
The incidents in this scenario serve to highlight the differences between the two main approaches described above. Firstly, for those who hold the view that data are not an “object” for the purposes of IHL, there is little difference between all three incidents. According to that view, because data is not an object, cyber operations directed against the data are not governed by the IHL rules regulating targeting, which include the definition of military objectives. Consequently, for those who hold this view, the deleting or corrupting any of the three datasets would be lawful—at least as regards the IHL rule of distinction. In other words, none of the operations conducted by armed group G in incidents 1–3 would amount to a violation of the rule of distinction. Instead, the legality of the manipulation of the data would be considered in light of other rules applying to military operations in general, such as the duty to take “constant care” to “spare the civilian population, civilians and civilian objects”.

On the second view, according to which data may be an “object” under IHL, the lawfulness of the relevant operations would have to be assessed with reference to the elements of the definition of military objectives. In this regard, the data targeted in incident 1 would likely qualify as a legitimate target of attack under IHL. This is because datasets stored in a military network and consisting of information on military assets belonging to the adversary are inherently military in nature. In addition, these datasets “contribute to the execution of the enemy’s operations or otherwise directly support the military activities of the enemy”. As such, they make an effective contribution to the adversary’s military action by their nature, fulfilling the first prong of the definition under Article 52(2) AP I.

By denying the government armed forces immediate access to the information about their own military assets, the insurgents impede the government’s military action, likely forcing State A’s armed forces to allocate resources towards the restoration of the lost information and away from the military operations. The destruction of the data would thus offer a definite military advantage “in the circumstances ruling at the time”, satisfying the second prong of the definition under Article 52(2) AP I.

In contrast, the data targeted by the cyber operation in incident 2 would likely be a civilian object and therefore not liable to attack in accordance with IHL. Datasets used and kept for strictly non-military purposes only do not meet the definition of military objective: they do not make an effective contribution to military action by nature, location, purpose or use. As civilian objects, they are therefore are protected from attack during armed conflict. As noted, this analysis fosters the protection of essential civilian datasets and, consequently, it aligns with the object and purpose of the relevant legal norms.

Incident 3 is the most complex of the analysed operations. Ordinarily, the activities of a civilian press agency, even if operated by the government in the context of an ongoing NIAC, do not contribute towards any belligerent’s military action. Exceptionally, specific media reports might effectively contribute to the enemy’s operational picture, and as such, depriving the enemy of them might offer a definite military advantage. Accordingly, in these exceptional situations, the data containing such reports would qualify as a legitimate military objective. However, the deletion of all data belonging to a press agency and its replacement with the insurgents’ propaganda would most likely go beyond such a narrow goal and therefore, the cyber operation would appear to be a case of unlawful targeting of a protected civilian object.

However, it is difficult to square this prima facie conclusion with the fact that States frequently engage in psychological operations of this kind. This may be perceived as a sign that the view interpreting data as an “object” under IHL is “over inclusive”. Nevertheless, a possible response to such criticism is that through the longstanding, general, and unopposed practice of States, a permissive norm of customary law has emerged, which specifically permits psychological operations and dissemination of propaganda directed at the civilian population, irrespective of the means through which such operations may be conducted. On the basis of this interpretation, the cyber operation against State A’s press agency in incident 3 qualifies as permitted under IHL, even if the starting point of the analysis is that data constitute an “object” for the purposes of IHL.

It must be stressed that the analysis above assumed that, in all three incidents, the data was deliberately targeted—i.e., its manipulation was the objective of the operation in question. However, there are at least two types of scenarios where the manipulation of data would be analysed differently, even if such manipulation was regarded as an “attack”. First, it may be that the target of the attack is not the data yet the operation foreseeably results in its incidental manipulation. Second, it may be that the operation in question is not a targeting operation because its objective is not the destruction or neutralization of a certain object but rather something else (for example, gaining access to a system or collecting intelligence about its weaknesses), yet the operation nevertheless foreseeably results in the incidental deletion or corruption of data. In both of these cases, whether the data qualifies as an “object” has no bearing on the permissibility of the attack under the rule of distinction. At the same time, if one takes the position that the data constitutes an “object”, the manipulation must be considered as collateral damage and will therefore have to be assessed and acted upon as the rules on precautions and proportionality require.

In sum, the law is unsettled as to the qualification of computer data under the targeting rules of IHL. Accordingly, States’ views aligning with one or the other of the approaches detailed above are needed in order to facilitate legal certainty in this area. In the meantime, the following table serves to highlight the points of difference between the two dominant interpretive approaches:

Checklist

 * Does the situation qualify as either an international armed conflict or a non-international armed conflict?
 * Are the rules of targeting applicable to cyber operations?
 * Is data considered to qualify as an “object” for the purposes of IHL?
 * Does the cyber operation target datasets constituting military objectives only?
 * Does the cyber operation target essential civilian datasets?
 * Does the cyber operation target non-essential civilian datasets?
 * If the cyber operation is not targeted at the data, yet it foreseeably results in its manipulation, does the manipulation accord with the rules on precautions and proportionality?

Bibliography and further reading

 * William H Boothby, The Law of Targeting (OUP 2012)
 * Yoram Dinstein (ed), The Conduct of Hostilities under the Law of International Armed Conflict (CUP 2016)
 * Heather A Harrison Dinniss, ‘The Nature of Objects: Targeting Networks and the Challenge of Defining Cyber Military Objectives’ (2015) 48 IsrLR 39
 * Noam Lubell, ‘Lawful Targets in Cyber Operations: Does the Principle of Distinction Apply?’ (2013) 89 Int’l L Studies 252
 * Kubo Mačák, ‘From Cyber Norms to Cyber Rules: Re-engaging States as Law-makers’ (2017) 30 LJIL 877
 * Kubo Mačák, ‘Military Objectives 2.0: The Case for Interpreting Computer Data as Objects under International Humanitarian Law’ (2015) 48 IsrLR 55
 * Noam Neuman, ‘Challenges in the Interpretation and Application of the Principle of Distinction During Ground Operations in Urban Areas’ (2018) 51 VJTL 807
 * Marco Roscini, Cyber Operations and the Use of Force in International Law (OUP 2014)
 * Yves Sandoz, Christophe Swinarski and Bruno Zimmermann (eds), Commentary on the Additional Protocols of 8 June 1977 to the Geneva Conventions of 12 August 1949 (ICRC 1987)
 * Michael N. Schmitt, ‘International Cyber Norms: Reflections on the Path Ahead’ (2018) 111 Netherlands’ Military Law Review 12
 * Michael N Schmitt, ‘The Notion of ‘Objects’ during Cyber Operations: A Riposte in Defence of Interpretive and Applicative Precision’ (2015) 48 IsrLR 81
 * Michael N. Schmitt and Sean Watts, ‘The Decline of International Humanitarian Law Opinio Juris and the Law of Cyber Warfare’ (2015) 50 TexIntlLJ 189
 * Silja Vöneky, ‘Analogy in International Law’, in R Wolfrum (ed), Max Planck Encyclopedia of Public International Law (OUP 2008)

Contributions

 * Scenario by: Taťána Jančárková & Kubo Mačák
 * Analysis by: Kubo Mačák
 * Reviewed by: Reviewer245, Reviewer715, and Reviewer354