Editor's note: This post by Sean Lawson provides context on cyber conflict, an area of interest at the nexus of national security and technology. - bg
Recently, Dr. Thomas Rid of the War Studies department of King's College in London published an article in the Journal of Strategic Studies titled, "Cyber War Will Not Take Place." Rid's essay relies upon a definition of war taken from the work of Carl von Clausewitz to assess whether cyber attacks can be accurately described as "stand-alone" acts of war. His conclusion is that we have yet to see any cyber attacks that, on their own, meet Clausewitz's definition. What's more, he predicts that we are unlikely to see stand-alone acts of cyber war in the future. Nonetheless, he does acknowledge that cyber threats are real and that various cyber tools and techniques are becoming increasingly important in international conflict, including those used for sabotage, espionage, and subversion.
At a time of increasing concern over prospective cyber threats, it is not surprising that Rid's essay has added fuel to the ongoing debate between cyber security proponents and the so-called "cyber skeptics." For example, cyber war expert, author, and CEO Jeffrey Carr has written a spirited response to Rid's essay. In this post, I argue that Carr's response misses a key component of Rid's argument, that the debate between Rid and Carr is exemplary of an emerging debate over the definition of "war" more generally, and that the complexities of cyber conflict demand that we move beyond the kind of binary thinking exhibited in this debate.
First, Carr provides three examples of cyber attacks that he says meet the Clausewitzian definition of war provided by Rid because all three are "lethal, instrumental, and political." His three examples:
- Kyrgyz Intelligence assassinates Gennady Pavlyuk. Kyrgyz intelligence cracked Pavlyuk's email account and used the information they obtained to lure him out of the country under false pretenses resulting in his murder.
- Mossad assassinates Mahmoud Al-Mabhouh. Israel's Mossad mounts an operation to assassinate Hamas leader Mahmoud Al-Mabhouh which includes infecting Al-Mabhouh's computer with a trojan horse virus.
- Iran's IRGC arrests 30 dissidents after cracking U.S. hosted webservers.
None of these are acts of war in the conventional sense of the term. These are 1) subterfuge in support of assassination, 2) espionage in support of assassination, and 3) espionage in support of political repression. In the first case, Kyrgyz intelligence supposedly assassinated one of its own citizens. That is not war as we typically understand it. In the third case, espionage was used to aid in carrying out an act of political repression. But neither of those acts by themselves (espionage nor arresting one's own citizens) are war. The only example that might be considered war is the second case. But even here, given the ongoing state of violent conflict between the Israelis and Palestinians, which has included many assassinations, it is hard to see this event as somehow distinct.
But more importantly, Carr's response misses a key component of Rid's argument, namely, that it was about whether stand-alone cyber attacks have been or will be acts of war. Not only is it questionable whether any of Carr's counter examples in their totality are "acts of war," it is clear that in none of them can the cyber attack components be seen as stand-alone acts of war. The cyber attacks in each example were not the direct causes of the ultimate outcomes. Email hacking did not directly kill Gennady Pavlyuk. The trojan horse did not kill Mahmoud Al-Mabhoud. Cracked servers did not directly arrest those Iranian protestors. All of those actions (assassinations and arrests of political dissidents) have occurred, do occur, and will continue to occur without the aid of cyber attacks. The use of cyber attack tools and techniques in support of them in these cases does not make them nor the use of cyber tools and techniques "acts of war."
Second, the debate between Thomas Rid and Jeffrey Carr is exemplary of an emerging debate that is less about the definition of "cyber war" and more about the definition of "war" in general. There is an emerging debate between expansionists and traditionalists. Expansionists argue that current definitions of "war," either from the classic theorists like Clausewitz or the law of war, are inadequate and should be expanded to include a wide range of acts that traditionally would not be considered war. The traditionalists argue that existing definitions of war are more than adequate, that while the practice of war might change (including weapons and tactics) the fundamental nature of war does not: it is still about damage, destruction, injury, or death inflicted for political purposes, usually by state actors.
In this instance, Carr makes an expansionist argument when he claims that "traditional thinking about warfare has been made obsolete by our dependence upon cyber-space-time." In a previous essay, he cited a NATO study of the legal lessons learned as a result of the 2008 cyber attacks against the country of Georgia. That report concluded that the cyber attacks, by themselves, did not count as "armed attack" (the legal term for what we colloquially call an "act of war") under current definitions in the law of war. In response, the authors proposed that "new approaches to traditional law of war principles need to be developed." Therefore, they advocated that the advent of "new bloodless types of warfare" like cyber war mean that "the definition of 'attack' should not be strictly connected with established meanings of death, injury, damage and destruction" (p. 30). Because the cyber attack on Georgia (and practically all other cyber attacks to date) do not come close to meeting traditional definitions of war from law of war, theorists like Clausewitz, or even common understandings of the term, the response has been to call for the redefinition of war itself to include a whole host of "bloodless" acts.
Of course, Rid is taking a traditionalist approach in this debate. He is arguing that the fundamental nature of war has not changed and is using the work of a widely-cited, well-respected classic theorist to support his argument. Others, such as Maj. Gen. Charles Dunlap, Jr. (ret) have also taken the traditionalist position. Dunlap has argued convincingly that the law of war definitions of "armed attack" are more than adequate for evaluating cyber attacks. In doing so, he is following closely the analysis provided by Michael Schmitt more than a decade ago, which is the foundation for the "effects-based" approach to determining when a cyber attack rises to the level of armed attack, when it is war. In the traditionalist view, there are no "bloodless" acts of war. Violence, death, destruction, damage, injury are required. Even then, not every act of this sort is armed attack or war (see examples 1 and 3 above).
My own views are more in line with those of the traditionalists than the expansionists. I believe that it is dangerous (for many reasons that I will not elaborate here) to expand definitions of armed attack and war. I merely wish to call attention to the fact that the debate over the definition of cyber war is becoming a debate over the definition of war in general. This is an important distinction. The outcomes of this debate will have profound impacts on the future of politics, economics, security, and individual liberties.
Finally, because the outcomes of this debate will be so important, it is all the more disappointing that issues of cyber security are so often framed in such binary terms. For example, because Rid does not accept that all malicious cyber activity is "war," Carr lumps him into the category of cyber war "skeptics." He made a similar move with the authors of a recent OECD report that claimed that cyber attacks do not have the ability to cause systemic shocks that are global in scope. Nonetheless, in each case, Rid and the authors of the OECD report make it clear that they take cyber threats seriously. They merely seek to be more realistic in their assessments of the impacts of cyber attacks and more precise in their categorization of the varying types of malicious actions in/through cyberspace. To be fair, though Rid's position is more nuanced than Carr admits in his response, Rid nonetheless invites Carr's application of a binary, proponent/skeptic categorization because his essay largely framed cyber war as a yes/no question.
Effectively addressing the complex challenges of cyber security in a globalized world demand that the public debate about cyber security move beyond such framings. Are cyber attacks war? There needs to be room for answers like, "Maybe, it depends" and "No, but there are still serious challenges." The truth will likely lie somewhere in the gray, muddled middle between yes and no, black and white. Engaging with the messy complexity of cybersecurity challenges is essential to ensuring that war remains a continuation of politics by other means instead of politics (and every other aspect of daily life) becoming a continuation of war by other means.
[Cross-posted from Forbes.com.]