Two Case Studies of Clandestine Operations, Attribution and Functional Immunity for Ordinary Crimes – EJIL: Talk!

In this post, I want to briefly analyze two recent high-profile developments. The first is the mass prisoner exchange (if that’s the right term) between Russia and Western states that happened some two weeks ago. That exchange involved the release of political prisoners by Russia, and several Russian secret agents by the West. One possible interpretation of that exchange is that Russia coerced Western states into releasing individuals legitimately convicted for serious criminal offences, essentially be using innocent people as hostages, in a manner that at least arguably violates not just their human rights, but also the prohibition of intervention (see my AJIL article on coercion, at IV.C, dealing with so-called hostage diplomacy as a way of conducting coercion-as-extortion). But the issue that concerns me in this is post is how, in doing the exchange, Russia acknowledged that some of the individuals concerned had previously acted on Russia’s behalf, i.e. that they were its agents, something that it had previously denied.

The second development is a story published yesterday in the Wall Street Journal regarding the 2022 sabotage operation against the Nord Stream pipelines in the Baltic Sea. The WSJ story claims that the sabotage was conducted by Ukrainian team, in an operation initially authorized by President Zelenskyy. The sabotage team is said to have been composed of Ukrainian soldiers and civilians with special expertise. The WSJ story alleges that, after the CIA learned about the operation, Ukraine was asked by the United States to call off the operation. President Zelenskyy issued such an order, but the commander in chief of Ukraine’s armed forces, Valerii Zaluzhnyi, basically went ahead with it anyway.

These two case studies are different in that the facts of the first are more or less beyond doubt, whereas the same cannot be said of the second. In particular, I have no idea whether the WSJ’s reporting is accurate, but I will assume for the sake of the argument that it is. As a legal matter, however, both case studies raise similar questions. The first is establishing the attribution of conduct as a matter of the law state responsibility. The second is whether individuals who act on behalf of the state enjoy – or not – some kind of functional immunity, as they are generally understood to. Both episodes are striking in how states initiate criminal proceedings against individuals who they believe acted on behalf of another state, yet for some reason functional immunity is not considered to be a bar to such proceedings – and I would really like to know why that is the case.

Attribution and the Nord Stream Operation

Assuming the facts alleged in the WSJ report are true, (1) the operation was conducted by a mix of Ukrainian soldiers and civilians who were not Ukrainian state organs; (2) it was done at the orders of General Zaluzhnyi, but (3) contrary to the orders of President Zelenskyy. This is an interesting fact-pattern, but one that the rules of attribution in international law, as authoritatively stated in the International Law Commission’s Articles on State Responsibility, can deal with reasonably easily.

Re (1), the conduct of Ukrainian soldiers is that of de jure state organs, and would be attributable to Ukraine under Article 4 of the ILC Articles on State Responsibility. The conduct of the contracted civilians would be attributable to Ukraine under Article 8 ASR – instructions, directions or control. The interesting point here is how the ultra vires rule in Article 7 of the ILC Articles would apply here, since one Ukrainian official ordered the operation, while his superior countermanded it, without the operators on the ground apparently being aware of this. Moreover, Article 7 normally applies to the ultra vires conduct of state organs, but not to the conduct of non-state actors under Article 8. In other words, the attribution rule in Article 8 does not apply if instructions given are exceeded.

My own sense of this is as follows – it was General Zaluzhnyi who acted ultra vires, because he disobeyed an order given to him by President Zelenskyy. But the saboteurs did not actually act ultra vires, since they received their orders from the general. In particular, the civilian contractors – whatever their exact role was – were at all times acting under the instructions of de jure organs of Ukraine, including General Zaluzhnyi and any of his subordinates, and they did exactly what was expected of them (assuming the WSJ reporting is true). Those instructions may themselves have been ultra vires, because of President Zelenskyy’s order, but that conduct – the exercise of control over the saboteurs by General Zaluzhnyi – clearly remained attributable to Ukraine. That means that all of the elements of the operation can be attributed to Ukraine, regardless of whether they were committed by Ukrainian soldiers or by contracted civilians, and regardless of President Zelenskyy’s order.

Attribution and Russian Spies and Assassins

Turning to the Russian-Western prisoner exchange, an interesting point here is how Russia has now officially acknowledged that some of the individuals concerned acted on Russia’s behalf, which it had previously denied. Thus, according to the Guardian,

The Kremlin has admitted that Vadim Krasikov, the assassin freed by Germany in a historic prisoner swap on Thursday, is a serving officer of the Federal Security Service of the Russian Federation (FSB), essentially an acknowledgment that his 2019 murder of a Chechen exile in Berlin was a state-ordered hit.

It also hinted that he was linked to Vladimir Putin’s personal guard.

“Krasikov is an FSB employee,” Putin’s spokesperson, Dmitry Peskov, told reporters, adding that he had “served with some of the people working in the president’s security detail”.

Similarly:

Peskov confirmed that Artem Dultsev and Anna Dultseva, who had been posing as an Argentine couple in Slovenia, were in fact Russian “illegals” – deep-cover spies who can spend decades abroad pretending to be foreigners. The couple’s two children, who had been taken into foster care when their parents were arrested in late 2022, travelled to Russia with them.

How does this post facto acknowledgement fit within the ILC attribution framework? Some might be tempted to invoke Article 11 ASR, under which ‘Conduct which is not attributable to a State under the preceding articles shall nevertheless be considered an act of that State under international law if and to the extent that the State acknowledges and adopts the conduct in question as its own.’ But this is not the relevant rule here. Rather, the conduct of Krasikov and the Dultsevs would be attributable to Russia under the organ rule in Article 4 ASR. At the time all these people acted, they were organs of the Russian Federation – as FSB or GRU employees or whatever, even if Russia publicly denied that they were such.

What Putin’s red carpet welcome of the spies and assassins, coupled with the statements of his spokesman, do, is provide incontrovertible evidence that the individuals concerned were, and remained, Russian state organs. Article 11 ASR does not apply simply because Article 4 does – it is not ex post facto acknowledgement and adoption that is the basis for attribution, but the fact that when they acted these individuals were Russian state organs, any attempts at obfuscation notwithstanding.

In short, this case study well illustrates how we need to keep a clear conceptual separation between the legal standard of attribution that may apply and the evidence that the requirements of the legal standard were met. Consider, for example, cyber operations that are run by state intelligence agencies or their militaries. The conduct of such cyber operators would be attributable to the state under the organ rule in Article 4 ASR. However, states will almost invariably deny that they engaged in some potentially unlawful cyber operations. The issue of evidence is whether, at the time the operation is conducted or subsequently, there is sufficient proof that state agents conducted the operation, despite the state’s denial. Obviously, such evidence can emerge after the fact. For instance, the Stuxnet malware operation can be attributed to Israel and the United States on the basis that their organs conducted this operation. The evidence here is, inter alia, that a showreel played at a retirement party for the head of the Israel Defence Forces showed their involvement with Stuxnet, and that a US general was prosecuted for disclosing US and Israeli execution of Stuxnet to a US reporter. That is simply evidence of attribution under the organ rule. Putin on the red carpet is really no different.

Functional Immunity for Clandestine Operations and Ordinary Crimes

This brings me to my final point. Much ink has been spilled in the past few decades, but also more recently in the context of the ILC’s work on the topic, on the question whether state official who commit core international crimes benefit from functional immunity. The idea of functional immunity (or immunity ratione materiae) is that any individual who acts on behalf of a state should be immune from civil suit or criminal prosecution for such acts, even after they are no longer a state official (unlike immunity ratione personae). And so, in their comments to the ILC, some states cast doubt on the ILC’s draft rule that functional immunity does not apply to international crimes, arguing that this is not necessarily the position under customary international law (see more here and here; the ILC itself was also split on the issue). As noted recently by Benjamin Meret, some of the states appeared to take contradictory positions between their comments to the ILC and their support for domestic prosecutions of Russian officials for crimes against international law in Ukraine – to that one could add a substantial number of universal jurisdiction cases worldwide against former state officials on matters other than Ukraine.

But what about our two case studies – what do they have to do with functional immunity? Well, in both of these situations the national authorities of one state made a determination that a crime was committed in a clandestine operation conducted by the officials of another state, yet – any functional immunity rule notwithstanding – they nonetheless initiated criminal proceedings against these individuals. And these crimes are ordinary crimes under domestic law, such as murder, sabotage or espionage, not crimes under international law.

So, for example, as is frequently the case in espionage cases, Slovenia put the two Russian ‘illegals’ – the Dultsevs – on criminal trial under Slovenian law, and duly convicted them before they could be exchanged. Slovenia did this even though it knew that the Dultsevs worked for Russian intelligence, either through its own investigative efforts, or indeed because ‘Russian officials had swiftly admitted in private that the pair were intelligence officers and discussions had started over a possible exchange.’ But, for some reason, functional immunity was not regarded as a bar to prosecution.

Similarly, Germany prosecuted Krasikov for murder, even though (at some point) it knew that he was a Russian state agent. When it comes to the Nord Stream sabotage, it was reported a few days ago that German authorities issued a European arrest warrant for a Ukrainian man named Volodymyr Z. – obviously, not that Volodymyr Z., but you just have to love the coincidence here – who was one of the divers in the sabotage operation. And again, Germany did so even though its authorities likely believe that the operation was conducted by Ukrainian state organs.

To these examples we could add a great many others, ranging from the Rainbow Warrior incident, to say the United States government filing indictments against hackers that the United States itself claims are employed by the Chinese military, for committing completely ordinary crimes, not international crimes – yet, somehow, functional immunity does not apply.

I think it’s fair to say that there is some inconsistency between this kind of state practice and the position that some of these states have taken with regard to functional immunity for international crimes. If a Chinese military hacker is not functionally immune for committing an ordinary cyber crime under US domestic law, then how could immunity exist for acts that constitute war crimes or crimes against humanity under international law?

Some arguments could be made, and have been made in the literature, to explain the application (or not) of functional immunity in these types of cases:

• Functional immunity only applies if the state on whose behalf these individuals act officially invokes that immunity, which it can’t do if it’s denying that these individuals have acted on its behalf. But what if a state does make such an official acknowledgement? France did acknowledge the Rainbow Warrior incident and did expressly rely on functional immunity, yet New Zealand nonetheless continued with the prosecution (for an ordinary crime). Similarly, had the UK been able to apprehend the perpetrators of the Salisbury nerve agent incident in 2018, I’m pretty sure they would have been prosecuted even if Russia expressly acknowledged that Team Novichok worked on its behalf. And again the core problem remains that the prosecuting state believes that the individuals concerned acted for another state.
• There is some kind of espionage exception, or perhaps a broader clandestine operation exception, from functional immunity – but why? And how could that be squared with positions taken by some states regarding vastly more serious international crimes?
• Functional immunity does not attach if state agents commit a crime on the territory of another state. But again, why? Isn’t the agent nonetheless acting on the state’s behalf? Are we really going to say that if FSB agents use Novichok in Salisbury, they enjoy no functional immunity, but if they poison Navalny in Russia, that they do? And how on Earth would this rationale even apply to the Nord Stream operation, which was conducted on the high seas and not within the territorial seas of Denmark or Germany?

To be clear, I have no idea what the right answer is here. The only point I’m making is that there is some inconsistency between, on the one hand, a high degree of angst expressed by some states regarding a rule that says that no functional immunity exists for genocide, crimes against humanity and war crimes, and, on the other hand, the genuinely casual way in which some of these same states have prosecuted foreign officials for purely ordinary, domestic crimes that pale in comparison to crimes under international law. It would be good if some of these states made public their views for why functional immunity does not apply to clandestine operations of the kind examined in this post.