State Responsibility in Crisis

Introducing the ‘State Responsibility in Crisis’ Symposium

On 30 April and 1 May 2026, a group of academics came together at a workshop hosted by the University of Bristol, with the kind support of a BA/Leverhulme Small Grant, to discuss a simple but pressing question: can the law of state responsibility still function as a meaningful framework of accountability in a globalised, multilateral order, or is the regime itself facing a crisis?

The contributions to this symposium reflect the papers discussed at the workshop and take up this question from doctrinal, theoretical, and socio-legal perspectives. They examine how the law of state responsibility operates in contemporary contexts and explore problems of attribution, causation in fact and law, shared and joint responsibility, omissions, due diligence, and the relevance of fault across different substantive fields. Collectively, they ask whether the conceptual tools of the Articles on Responsibility of States for Internationally Wrongful Acts (ARSIWA)  can be reinterpreted and extended to meet present demands, or whether a more fundamental rethinking of responsibility in international law is required.

At stake is more than doctrinal coherence. It is whether international law can continue to provide meaningful frameworks of accountability in a world where power is exercised through networks, systems, and collaborations that defy the bilateral assumptions on which the law of State responsibility was built.

The Law of State Responsibility in the Modern Context

The law of state responsibility was built for a world that no longer exists. Its architecture – expressed in the ARSIWA – emerged from practice shaped by bilateral disputes. One State injured another; responsibility could be traced through relatively linear pathways of attribution, breach, and causation. The rules assumed identifiable actors, discrete wrongful acts, and harms that could be connected to conduct through familiar causal reasoning.

Today, the problems that dominate international law rarely fit this model. Armed conflicts involve coalitions, proxy actors, private contractors, cyber operations, and AI-supported systems. Climate change produces cumulative, dispersed, and temporally extended harm. Displacement and refugee protection unfold through externalised borders, third-State cooperation, and delegated control. States increasingly act in concert with other States, international organisations, and private entities, creating dense webs of interaction that blur the lines between those who act, those who assist, and those who enable.

In this setting, the secondary rules of responsibility are being asked to do work for which they were not originally designed. The workshop keynote, delivered by Professor Violeta Moreno-Lax, provided a conceptual lens for this inquiry. She diagnosed what she calls three interlocking crises of the international responsibility regime – concerning actors, conduct, and harm – and urged a move toward an integrated understanding of responsibility grounded in the rule-of-law intuition that wherever power is exercised, accountability must follow. This symposium takes that diagnosis as a starting point.

From Bilateral Injury to Multilateral Governance

The ARSIWA’s structure reflects a reciprocal model of inter-state claims. Much of the practice that informed the International Law Commission’s work arose from situations in which one State directly injured another. Responsibility could be conceptualised as the restoration of balance between two legal equals.

Contemporary international law is organised differently. Many of the issues that international law regulates – armed conflict involving multiple participants, global environmental degradation, transnational migration, global supply chains, digital infrastructures – are inherently multilateral. They concern global commons, community interests, and shared regulatory spaces.

Yet these are precisely the areas where secondary rules of responsibility face their greatest strain. Attributing a breach becomes uncertain when conduct is dispersed across multiple actors. Causation, necessary to establish that a breach of international law has occurred and integral to establishing reparations, becomes complex when harm is cumulative or structurally produced. These core processes in determining responsibility struggle to operate in a multilateral, multi-actor context. Further, the distinction between direct responsibility and complicity blurs when States cooperate in shared governance arrangements. The relevance of fault becomes difficult to assess where duties of due diligence, prevention, and cooperation dominate.

The traditional state responsibility model struggles to capture “actor-less” harm, multi-actor collaboration, omissions in complex regulatory environments, and aggregate or structural forms of injury. The problem is not simply doctrinal difficulty, but a mismatch between the regime’s conceptual foundations and the realities of contemporary governance.

Three specific contexts prompted my consideration of this shift towards a multilateral, multi-actor, multi-causal world in which the law seeks to hold states responsible for breaches of international law: armed conflict, climate change, and displacement. This symposium explores how these contexts – and others – impact the law on state responsibility, highlighting the challenges they present and potential solutions.

Armed Conflict and the Problem of Shared Action

Modern conflicts are rarely bilateral territorial disputes. They are often internationalised, asymmetric, and urban, involving States, non-state armed groups, supporting States, arms exporters, peacekeeping missions, and private contractors. Military operations are coordinated across alliances; assistance is provided without formal party status; logistics, intelligence, and technology are outsourced.

These constellations obscure clear lines of attribution. When harm results from joint operations, intelligence sharing, arms transfers, or logistical support, how should responsibility be allocated? Are assisting States merely ancillary actors, or do they bear direct responsibility for foreseeable consequences? At what point does cooperation in a conflict environment amount to aiding or assisting internationally wrongful acts?

The ARSIWA and the Articles on the Responsibility of International Organizations (ARIO) presuppose that responsibility attaches to States and international organisations. Yet contemporary governance involves powerful non-state entities—corporations, funds, private organisations – whose actions may be harmful without being easily characterised as internationally wrongful.

The traditional response has been to attempt to attribute their conduct to States through doctrines of effective control or delegation. But this misses the point. Some actors exercise real power and produce real harm independently of States, yet fall outside the framework of international legal personality and obligation. In such cases, the difficulty lies not only with secondary rules of responsibility, but with the primary rules that determine who can bear obligations at all. The limits of responsibility doctrine expose deeper questions about subjecthood in international law.

Digital operations and AI-supported systems further complicate matters. Cyber warfare, automated targeting, and “human-in-the-loop” systems challenge traditional assumptions about control and decision-making. If human oversight becomes nominal due to machine bias, how should causation and attribution be understood?

The law of responsibility assumes that wrongful conduct can be traced to a primary actor exercising effective control. But in networked warfare, power is distributed, and harmful outcomes may be the product of many contributions that are individually lawful but collectively devastating.

Climate Change and Complex Causation

Climate change presents a different form of stress test for the responsibility regime and represents a wider challenge it faces. Harm is cumulative, geographically dispersed, and temporally extended. No single State’s emissions can be said to “cause” a particular hurricane, drought, or sea-level rise in the traditional sense. Yet scientific knowledge allows us to understand these harms as the foreseeable outcome of aggregated conduct.

As such, many harms arise not from a single wrongful act, but from the aggregation of individually lawful decisions, systemic governance structures, or non-agentic forces such as environmental degradation. Linear causation struggles to make such harm legally legible. Courts are increasingly asked to translate complex scientific reasoning into legal standards of responsibility. Concepts such as common but differentiated responsibilities, loss and damage, and precaution interact uneasily with ARSIWA’s traditional structure.

Where States know – or ought to know – about the foreseeable consequences of their conduct, responsibility may arise not from direct causation of a specific harm, but from participation in a risk-generating system. Yet this approach remains fragmented across jurisdictions and regimes, with no coherent, principled test.

This raises the possibility that certain forms of structural violence may remain beyond the reach of traditional responsibility doctrine unless new approaches to causation and harm are developed.

The difficulty here is not merely evidentiary. It is conceptual: can the law of state responsibility adapt to forms of harm that are structural, probabilistic, and cumulative?

Displacement, Externalisation, and “Contactless Control”

The global displacement crisis reveals another dimension of the problem. While displacement has reached historic highs, many States resist meaningful responsibility-sharing. Instead, migration control is externalised through agreements with third States, international organisations, and private actors. Borders are projected outward; obligations are mediated through layers of delegation.

Traditional doctrines tend to place primary responsibility on the country of origin or occupying force. Yet States that fund, equip, train, or coordinate migration control practices often avoid legal responsibility because they do not exercise direct territorial or physical control. Moreno-Lax has described this as “contactless control”. Power is exercised at a distance, but human rights courts often treat jurisdiction as a threshold that is not met in such cases. Jurisdiction, attribution, and causation become intertwined in ways that can create protection vacuums. Harm that would be unlawful if committed at home becomes legally invisible when pursued abroad through intermediaries.

Here again, the issue is not merely doctrinal nuance. It is whether the responsibility regime can account for contemporary forms of governance that deliberately create distance between power and its effects.

Can the Law on State Responsibility Respond to These Challenges?

Responding to these issues requires rethinking how attribution, causation, conduct, and harm interact, and how the general secondary rules relate to specialised primary regimes across international law. It may also require confronting the limits of existing primary rules concerning who can bear obligations in international law.

The contributions to this symposium take up this challenge. They explore how the law of state responsibility can be reinterpreted, supplemented, or reimagined in light of contemporary crises, examining issues of attribution, causation in fact and law, shared and joint responsibility, and the role of fault across different substantive fields of international law. Collectively, they ask whether the existing regime can stretch to meet present demands – or whether a more fundamental reconfiguration of how we understand responsibility in international law is required. As the UNGA Sixth Committee convened this October to consider the International Law Commission’s 2026 report, which considered several matters relating to state responsibility, we hope this symposium will offer valuable new perspectives to inform future UN discussions.

Overview

The symposium will begin with a post by Enenu Okwori on the nature of obligations and breaches of international law in cyberspace. Starting from the obligation to prevent transboundary harm, Okwori argues that states should be expected to monitor their cyber networks and infrastructure as a minimum due diligence standard, since early detection is more effective and cheaper than responding after an attack. Because due diligence is an obligation of conduct that varies with each state’s capacity, this baseline would make accountability clearer. States unable to meet it should seek cooperation or assistance from others, failing which they risk breach and countermeasures from victim states.

The symposium then turns to non-state actors and questions of attribution. In her post, Scarlett McArdle will argue that the international law of responsibility remains state-centric, leaving a growing gap as non-state actors, such as philanthropic foundations like the Gates Foundation, gain major influence in global health through collective action without legal personality or accountability. She contends that existing concepts like “non-state actor” and “participants” fail to capture how these actors can bear legal obligations. The resulting mismatch, she argues, threatens the legitimacy of the international legal system itself. Elisa Ruozzi will then examine the Houthis, who are simultaneously insurgents in Yemen’s civil war and attackers of international shipping. She shows how they resist classification: they fit poorly as pirates (given the “private ends” and “two-ships” requirements) or insurgents, face hurdles in qualifying as a non-international armed group under International Humanitarian Law, and are hard to link to Iran under ARSIWA’s attribution rules. These gaps sit uneasily with the legal consequences the Houthis nonetheless face, such as naval operations justified by self-defence and Security Council sanctions. Ruozzi sees in this an asymmetry in how international law treats non-state actors and a need for new legal tools.

Daisy Peterson’s piece then shifts to examine whether the conduct of incorporeal entities can be attributed. She questions whether the conduct of sovereign wealth funds, such as investments linked to human rights abuses, can be attributed to the State. She looks at three structural models. For autonomous entities, attribution under Articles 5 and 8 is difficult and fact dependent. For funds within a central bank, Article 4 organ status or a failure to supervise may ground responsibility. For funds under a Ministry of Finance, attribution is straightforward if the Ministry directs the decision but unlikely if management is outsourced to private asset managers. While the customary rules can in principle apply, structural design and a lack of transparency leave real attribution gaps, allowing harmful conduct, particularly by funds of authoritarian regimes, to slip through. Felix Butz’s contribution will continue to challenge modes of attribution for novel conduct, introducing the notion of “necropolitical structural violence”: diffuse, deadly harm produced through social structures that an actor can significantly influence and knows to be lethal. Butz argues that courts, especially in environmental and climate cases, have already begun to hold states responsible for it. Three shifts support this: accepting statistical knowledge of risk instead of identified victims, focusing on a state’s influence over structures rather than control over specific actors, and favouring structural remedies like cessation over individual compensation that depends on causation.

The symposium then moves to causation, a thread running through the following three posts. Tamás Molnár argues that causation, which ARSIWA treats as a settled criterion, is in fact constructed case by case. He shows this through the inconsistent tests (directness, proximity, foreseeability) that the ICJ, ECtHR, IACtHR, and CJEU apply to reparation claims for the unlawful expulsion of non-nationals. A foreseeability standard tailored to the primary obligation breached (following Pusztai) would, he concludes, offer a principled middle path between under-compensating individuals and drifting into punitive damages. Michael Oghenetega Abu will then consider climate-displaced children. He argues that responsibility should be assessed not only by whether a state caused the climate event, but also by whether timely, child-sensitive preventive measures could have interrupted the foreseeable path from climate risk to displacement and further rights violations. He acknowledges that courts remain cautious about linking such omissions to specific harms. A credible claim therefore needs an applicable obligation, a foreseeable risk, a reasonable preventive measure and a sufficiently plausible causal link, and how courts should decide when that link is strong enough remains open. Abdulkadir Gülçür will then show that investment tribunals routinely reduce awards for an investor’s contributory fault by round-number percentages, as in Occidental v. Ecuador and Yukos (25% each), without explaining how the misconduct caused the loss. They rely on broad discretion and old precedents like MTD v. Chile that annulment committees are reluctant to disturb. Since treaty or ARSIWA reform is impractical, Gülçür calls for tribunals to support reductions with concrete economic evidence, for annulment committees to treat unreasoned apportionment as a failure to state reasons under ICSID Article 52(1)(e), and for UNCITRAL or ICSID to issue soft-law guidelines on causation and contributory fault.

Finally, Antal Berkes will turn to the consequences of serious breaches for third States. He argues that the EU’s Forced Labour Regulation can give practical effect to Article 41(2) ARSIWA, which requires states not to recognise or assist situations created by serious breaches of jus cogens. In his reading, the Regulation targets serious modern slavery violations, identifies “situations” through its risk-based prioritisation of systemic, severe cases, and treats forced-labour products as their continuing consequence. It also requires Member States to adopt measures that substantially replicate non-recognition (banning and withdrawing such products) and non-assistance (supply-chain due diligence, investigations and inspections). Although the Regulation only applies from December 2027 and there is no evidence the EU sees it as implementing Article 41(2), Berkes concludes that it shows how legislation and corporate due diligence measures can help fulfil these obligations if applied consistently. Sissy Katsoni and Andreja Friškovec will then conclude the symposium with their reflections on the symposium contributions.