The international law governing national security and war is plagued by several critical ambiguities. When can states lawfully resort to armed force? What are the constraints on lethal military operations that threaten civilians? How certain must states be that a particular target is lawful? How does international security law apply to new technological domains such as cyber and artificial intelligence (AI)? When does international humanitarian law (IHL) apply at all? How does the application of these rules implicate the application of international human rights law (IHRL)? When are states responsible for illegalities committed by their allies? These questions and others were addressed earlier this month at a conference held at the U.S. Naval War College (USNWC) — The Fog of Law: Operating in the Grey Zones of International Law. The workshop, co-sponsored by Just Security, New York University School of Law, the Stockton Center for the Study of International Law at the USNWC, and the International Committee of the Red Cross (ICRC,) brought together a diverse group of experts including academics, military and other government lawyers, representatives of human rights groups, and the representatives of the ICRC. The conference’s goal was to identify, explain, and assess competing positions on critical issues with a view toward better understanding the “gray zones” of international security law.

Over the past month, several articles on Just Security, authored by conference session leaders, previewed the event by highlighting areas of disagreement in important issue areas. In the coming year, the session leaders will write academic papers analyzing the key debates of each issue. These papers will be published in the Stockton Center’s International Law Studies. This article provides a brief review of the workshop, focusing on ambiguity’s unique importance in both hindering and facilitating international law.

Important Debates in International Security Law: Case Studies in Ambiguity

The workshop addressed seven issue areas:

  • Sovereignty and cyberspace;
  • the legality of the use of force (jus ad bellum;)
  • the geographic and temporal scope of armed conflict;
  • the use of proxies;
  • targeting law;
  • human rights in armed conflict;
  • and AI.

Each session explored in detail a small set of representative ambiguities. The issue is not that international law is so abstract and under-specified that lawyers, policymakers, and operators are left to guess at what it requires. The workshop discussions did not suggest that international security law suffers from grievous ambiguities of that sort. Rather, the ambiguities discussed are points of law in which there has been persistent disagreement about interpretation, application, or status of important rules. The zone of reasonable disagreement is quite narrow on some issues and broader on others. Ambiguities in international security law are particularly important because the stakes are so high in these domains. Moreover, cooperation in addressing these ambiguities is often difficult to achieve through ordinary international lawmaking processes.

Prof. Michael Schmitt led the session on the state sovereignty and non-intervention in cyberspace. The discussion centered on whether respect for the sovereignty of states is a primary rule of international law — or whether it is simply a principle from which specific international legal rules may be fashioned. This in turn determines whether existing international law prohibits cyber intrusions that do not meet international law’s standards for an illegal intervention or use of force. If sovereignty is a primary rule, then cyber intrusions interfering with the target state’s sovereign prerogatives constitute a violation of existing international law. If sovereignty is, on the other hand, simply a general principle underlying concrete legal prohibitions, then current international law does not prohibit these smaller scale cyber actions. Both approaches enjoyed support from a number of participants. As reported and analyzed in Just Security, the UK Attorney General has only days ago weighed in on this debate — claiming that sovereignty is not a primary rule governing cyber operations.

The international law regulating when states may use force — jus ad bellum — was also discussed. Prof. Michael Glennon led the session — highlighting a number of ambiguities pertaining to the scope and content of the UN Charter use of force rules. These ambiguities include: how best to define a use of force; whether states may use force in self-defense against non-state actors; whether states may engage in in preemptive self-defense; and whether humanitarian intervention not authorized by the UN Security Council is consistent with international law in any circumstances. The most fundamental question raised was whether the pattern of systematic non-compliance with UN Charter use of force rules alters the legal status of those rules. Perhaps these rules are no longer valid because they have fallen into desuetude.

Prof. Jann Kleffner of the Swedish National Defence College led a discussion on the temporal and geographic scope of armed conflict. The issue was, of course, when and where international humanitarian law (IHL) applies. The discussion focused on how best to interpret Common Articles 2 and 3 of the 1949 Geneva Conventions and the applicability provisions of the two 1977 Protocols. Although there was a consensus on the definition of an armed conflict, meaningful disagreement emerged on the application of the concept in practice. Central problems included: how to define non-international armed conflict; how to define the end of hostilities; and whether there are meaningful geographic limitations on the applicability of IHL.

International rules governing when states are responsible for illegal acts committed by other states or non-state groups were also discussed. Prof. Ryan Goodman led a discussion that canvassed a range of legal constraints potentially applicable to the use of proxy forces. One difficulty here is whether states only incur responsibility if they intentionally promote or facilitate the wrongful acts of proxies. Another difficulty concerns whether and how the rules differ with respect to state proxies and non-state proxies. This session also explored state obligations to take affirmative steps to ensure respect for international law by other actors — including, in some circumstances, their proxy forces. Debate here focused on the meaning of Common Article 1 of the 1949 Geneva Conventions and Article 41 of the International Law Commission’s Draft Rules on State Responsibility. There was some disagreement about the scope and content of these obligations.

Prof. Adil Ahmad Haque led the discussion on ambiguities in the international law governing the conduct of hostilities. The session primarily concerned two persistent ambiguities in IHL’s law governing bombardment targeting. The first concerns how best to interpret the so-called rule of doubt concerning situations where it is unclear whether potential targets are civilian or military. Additional Protocol I requires that, in case of doubt, potential targets be presumed as civilian in character. It is unclear, however, what level of doubt triggers the applicability of the rule. This discussion led to a more fundamental debate about the appropriate standard of proof for establishing the lawfulness of a specific target. The second question discussed was the role uncertainty in a target’s status plays in the proportionality inquiry required by IHL. Participants advanced a range of views on these questions.

The workshop also addressed the complex relationship between IHL and international human rights law (IHRL.) Prof. Monica Hakimi of Michigan Law School led the discussion. She emphasized the many difficulties associated with claiming either that IHRL does not apply at all in armed conflict or that IRHL applies fully in the same way it does outside conflict. Discussion centered on the degree to which IHL displaces IHRL during armed conflict and how the two areas of law might be reconciled. Participants expressed diverging views on the extent to which IHRL is relevant and the extent to which the commitments of these two legal regimes might be harmonized.

Prof. Kenneth Anderson led the final session on AI — assessing whether developments in AI will fundamentally change the security environment, and whether these changes will strain prevailing international security law. Discussion centered on whether AI weapon systems and other AI-driven security tactics require a rethinking of traditional IHL principles. One area of particular concern was whether AI systems will be designed in a manner that impedes transparency and accountability for unlawful actions or mistakes. Participants disagreed on the extent to which AI-specific rules are advisable or required.

The Forms of Ambiguity in International Law

The workshop offers several insights on the form and function of ambiguity in international law. The “gray zone” case studies help illustrate several distinct types of ambiguity with distinct sources and differing costs and benefits.

Several of the issue areas reflect ambiguity in the interpretation or application of a settled rule. Examples include the debates about the interpretation and application of Common Articles 1, 2, and 3 of the 1949 Geneva Conventions. Another important source of ambiguity concerns how well-settled international legal rules apply to new technological developments such as cyber and AI. Other of the issue areas illustrate a deeper ambiguity in the status of a rule or principle. The most obvious example here is the debate about whether sovereignty is a primary rule of international law. Other examples include debates about the status of specific rules of state responsibility, the status of the rule of doubt and other precautions requirements for states not party to the Additional Protocols, and even debates about the continuing validity of the UN Charter’s use of force rules. Finally, some issue areas illustrate an even deeper ambiguity about which of multiple, competing, valid rule sets govern a specific situation. The most important example is the debate about the structural relationship between IHL and IHRL.

Some of these ambiguities are the unavoidable byproduct of a radical change in foundational assumptions about the technological, political, and economic parameters of the regulated domain. This kind of driver explains the numerous ambiguities surrounding international legal rules governing cyberspace and AI.

Other ambiguities are intentional. Intentional ambiguities might reflect a preference for flexible standards over rigid rules. One clear instance of this approach is the “armed conflict” trigger for the application of IHL. Intentional ambiguity might also represent a decision to accord substantial discretion to national-level authorities. A good example is the feasible precautions requirement in IHL targeting law. Finally, intentional ambiguity might simply reflect a lack of meaningful consensus about the optimal content of the rule. This is one way to understand the debates about the granular requirements of state responsibility and use of force. Intentional ambiguities will prove particularly difficult to resolve since the relevant law reflects a preference for ambiguity. These types of ambiguity are unlikely to be addressed unless the regime has failed to function as designed.

III. The Functions of Ambiguity in International Law

The workshop also illustrated some of the costs and benefits of ambiguity in international law. The discussions made clear that a thorough accounting of these costs and benefits for each issue area is necessary to identify the optimal level of legal uncertainty.

The costs of legal ambiguity are, of course, familiar. Ambiguity generates substantial tactical and operational uncertainty. It can also constitute an impediment to cooperation in addressing security threats — particularly if potential allies embrace different approaches to some critical legal issue. On a retail level, legal ambiguity can lead to systematic over-protection or under-protection of the relevant rights and interests.

Despite these obvious costs. several non-trivial benefits of legal ambiguity should also be noted. Ambiguity expands national-level freedom of action and discretion, and, more generally, it is a mechanism for decentralization of the international legal order. In a legal context characterized by radical asymmetries, the flexibility that arises from ambiguity might facilitate the participation of law-regarding states otherwise concerned about absorbing a comparative disadvantage vis-a-vis self-dealing, law-disregarding states. Ambiguity is preferable where the international community is unsure of the optimal level of regulation. On this view, substantial legal ambiguity might be preferred in the international legal regulation of cyber operations and the militarization of AI. Legal ambiguity also encourages a kind of experimentation. States will pursue divergent legal and policy strategies, all of which might be consistent with the under-determinative rule set. Finally, ambiguity might help promote compliance with IHRL and IHL because it might encourage greater levels of protection as a matter of policy. In other words, uncertainty about where lines are drawn might encourage over-compliance in risk-averse actors.

Mechanisms for Identifying and Resolving Ambiguities: Pathways to clearing the “Fog of Law”

 The workshop indirectly suggested several mechanisms for resolving legal ambiguities. The discussion also highlighted some of the comparative advantages and disadvantages of relying on these mechanisms. I mention them here briefly to spur further reflection on what approach might be most appropriate in specific contexts.

Most obviously, states could make new law to resolve any extant legal ambiguity. Options include express agreement via treaty or implied agreement via the emergence of customary international law. Multilateral treaties would resolve legal ambiguities most decisively. Such agreements, however, are elusive. High transaction costs and various persistent asymmetries in the international order likely preclude deep cooperation. Bilateral treaties might offer more promise — though such agreements would only resolve legal ambiguities between the treaty partners as a formal matter. Customary international law arising out of the iterated interactions between states could help address ambiguities over time. This approach is, however, decentralized and unpredictable. Persistent ambiguity is to be expected on this approach because specific customary law is unlikely to emerge quickly in several of these domains. In addition, state practice might not converge in some of these issue areas. A lack of transparency about state practice in security matters might also preclude meaningful clarification of customary international law ambiguities.

Another possibility is that states could rely on neutral, third party dispute resolution to address legal ambiguities. This approach has the advantage of low transaction costs, but states must surrender control of the direction of legal development. In addition, third-party dispute resolution systemically narrows legal uncertainty only if the views adopted by these bodies converge. Divergent international jurisprudence would only intensify the “fog of law.”

States and other relevant actors might rely on soft law processes to address legal ambiguities. The distinguishing characteristic of soft law approaches is that they do not give rise to binding legal obligations. The great advantage of this approach is that relevant stakeholders can assess and negotiate the preferred path forward without incurring much of the political cost associated with binding legal obligation. Prominent recent examples are the Tallinn Manual on the International Law Governing Cyber Operations and the ICRC Interpretive Guidance on the Notion of Direct Participation in Hostilities in IHL.