Lukas Schaupp (European University Institute) 

Going into 2026, few people will have had on their bingo cards that a modest European troop deployment in the Arctic could reignite the most serious transatlantic tensions in decades. Yet that is where things stand after the deployment of small military contingents from several European countries to Greenland prompted another tariff threat from U.S. President Donald Trump. Interpreting the move as a hostile signal, the U.S. president threatened to escalate tariffs until Denmark agreed to allow the United States to acquire Greenland, a proposition both Nuuk and Copenhagen have flatly rejected. 

The episode has fuelled debate about everything from Arctic security and NATO alliance politics to a U.S. president’s fascination with acquiring very large pieces of real estate and, crucially, whether Europe’s strategy of appeasing Trump has now officially failed. In this post, I focus on that latter question. Judging by recent reactions in Brussels and several EU capitals, patience appears to be wearing thin. This is most visibly reflected in the growing calls to deploy the EU’s Anti-Coercion Instrument (ACI), including from French President Emmanuel Macron and trade heavyweights in the European Parliament, such as Bernd Lange.

The EU’s Anti-Coercion Instrument is often described as the Union’s “trade bazooka” or even its “nuclear option”. Too often, however, these labels are used without much consideration of how the ACI actually works, or how it could be applied in a concrete situation like the current standoff. Having worked on the instrument quite a bit over the past few years (including in earlier work, available here, here, and here), I will therefore try to move beyond the rhetoric and offer a practical guide to the ACI’s procedure, as well as to what it can – and cannot – realistically achieve.

Like any good piece of EU legislation, the ACI (Regulation (EU) 2023/2675) starts, under Article 4, with diligent examination of the third country conduct, launched either ex officio by the Commission or at the request of a Member State (i.e. France or Denmark in this instance). The examination turns on whether the third country’s conduct qualifies as economic coercion under the Regulation, that is, whether it involves the application or threat of a measure relating to trade or investment to influence the sovereign choices of the Union or a Member State (Art. 2(1) ACI). Applied to the Greenland episode, the answer appears relatively straightforward. First, a threatened tariff increase is plainly a measure affecting trade. Second, the threat is explicitly and causally linked to a specific act by several Member States – the deployment of a small number of soldiers to Greenland. 

On its face, this looks like a textbook case of coercion as defined by the Regulation. What is less clear-cut is timing. The Regulation requires the Commission to act “expeditiously” and suggests that the examination should normally not exceed four months (Art. 4(2) ACI). That formulation leaves considerable room for discretion, not just for the Commission, but also for Member States to signal their preferences behind the scenes as the process unfolds. This flexibility need not be a weakness. In practice, even the credible prospect of an ACI investigation can already have effects, by signalling resolve and activating domestic actors in the U.S. who stand to lose from escalation. 

Crucially, this logic of the ACI also cuts the other way. This is where it gets more technical. Once the Commission concludes that economic coercion is present, it cannot make that determination on its own. Instead, it must submit a proposal to the Council of the European Union, which then has to actively adopt the implementing act by qualified majority (Art. 5(1) ACI). In practical terms, this means that at least 55 per cent of Member States, representing 65 per cent of the EU’s population, must support the determination within a period of around ten weeks (Art. 5(6) ACI), another reminder that timing under the ACI is anything but precise. This design choice matters. Rather than allowing the Commission to adopt the implementing act, unless a majority of Member States blocks it, as proposed initially, the ACI requires a positive coalition in favour of escalation. As a result, the determination stage becomes a natural focal point for domestic opposition, especially from export-oriented Member States concerned about further escalation, with Germany once again a prominent example. This dynamic goes a long way towards explaining why the ACI has not yet been activated, even in seemingly clear cases of economic coercion. 

Let us assume that the threat of hostile annexation of a Member State’s territory is sufficient to persuade the necessary majority to make a formal determination. Once that hurdle is cleared, the Commission is required to make a further attempt to engage with the third country, signalling – once more – that the EU would much rather avoid using the ACI at all (Art. 6 ACI). Only after this final diplomatic step can “response measures”, as the legal text calls them, be considered. At this stage, the Commission may adopt response measures by implementing act (this time reverting to the default voting logic) provided that three cumulative conditions are met: (1) the economic coercion has not ceased and injuries have not been repaired, (2) action is necessary to protect the interests of the Union or its Member States, and (3) the measures are considered to be in the overall Union interest (Art. 8(1) ACI). 

Once this stage is reached, the politics of selecting a response begin. The Commission may choose from a broad menu of possible response measures listed in Annex I of the Regulation. These range from familiar trade tools, such as tariffs on goods, to far more far-reaching options that cover the full scope of the EU’s trade competence, including trade-related intellectual property rights, export restrictions and, crucially in a U.S. context, trade in services. This latter category is where the ACI’s real bite arguably lies. Measures targeting digital services supplied predominantly by large U.S. tech firms would directly exploit a structural asymmetry in transatlantic trade. That said, moving from theory to practice is far from straightforward. Cutting off market access for service providers would be legally possible but politically difficult, given the lack of viable domestic alternatives. Taxing revenues generated by digital services appears more plausible, but would be technically complex and not implemented under the ACI. By contrast, imposing tariffs on goods, particularly if carefully calibrated to hit politically salient constituencies in the U.S., remains the most immediately available option. Yet, such tariffs can be imposed without the ACI, and given that previously announced retaliatory tariffs on U.S. imports worth €93 billion are merely paused, they could be reactivated far more quickly than any ACI-based response.

What does all this mean for the present question of whether the Greenland threats should trigger an ACI response? Triggering an ACI investigation should be almost a no-brainer. As I have tried to summarise, the way from the first official examinations to the actual imposition of measures remains purposefully long. Even at this stage, the ACI is designed to keep the door to de-escalation open almost until the very last moment, as response measures are normally subject to a deferred date of application (up to three months), during which the third country is formally notified, invited to negotiate, and given an opportunity to back down (Art. 8 ACI). Even if the U.S. threats are turned into actual coercion, demanding a response under the ACI, response measures are selected as part of a balancing exercise. Here, the Commission must design proportionate measures that do not exceed the injury suffered and that, prioritise effectiveness in ending the coercion while minimising collateral damage to the EU (Art. 11 ACI), an exercise that is far from straightforward. 

These constraints help explain why the ACI’s main value may lie less in the response measures it ultimately delivers than in the threat of escalation it is able to signal. The so-called TACO logic (“Trump Always Chickens Out”) ultimately rests on this idea that a credible threat of retaliation can be enough to induce restraint. China has shown this in other contexts, most notably through its use of export controls. The ACI can certainly contribute to this signalling function, even though the prospect of actual retaliation remains in practice far less clear than what the “bazooka” rhetoric suggests (a point best not overemphasised to our American friends). Much of this has to do with the technical and political complexity of regulating digital trade, where the EU’s most powerful options are also the hardest to deploy. Other tools, such as export restrictions, do have a huge coercive potential, but considerations of how to leverage these are just starting. If the U.S. were to destroy NATO by annexing Greenland, this might be the moment where the more destructive options available under the ACI become viable. Until then, these seem unlikely, as Europe continuous efforts to keep Washington engaged show.