High Stakes Promises: The Challenge of Security Guarantees for Ukraine

Historically, the term “security guarantee” has been used to indicate commitments by countries to militarily protect and come to the defence of other countries, although in some cases they have also simply meant a promise not to themselves attack those countries. Security guarantees are often, but not always, memorialised in legally binding treaties. It is their binding nature that distinguishes them from “security assurances”, which are non-binding political pledges. Although in reality neither guarantees nor assurances are enforceable, the former are widely understood to be higher forms of commitment. Countries that enter into a mutual defence treaty that includes reciprocal security guarantees are often referred to as “treaty allies”. Alliances undergirded by security guarantees create a form of deterrence. For example, any adversary that might attack a NATO member or other treaty ally of the United States knows that it risks a direct confrontation with the world’s most powerful military.

Among the most prominent treaty alliances are the 32-member North Atlantic Treaty Organization (NATO), in which the U.S. is the most powerful member, and the six-member Collective Security Treaty Organization (CSTO), in which Russia plays that role and the European Union, whose Treaty has a mutual defence clause. The U.S. participates in other multilateral alliances and has bilateral defence treaties with Japan, the Republic of Korea and the Philippines. Washington has also offered countries non-binding assurances. Most recently, President Donald Trump issued an executive order that pledges to defend Qatar in language that looks much like a guarantee, though under U.S. law such guarantees are generally understood to require Senate approval, which the order does not have. Moscow has signed mutual defence agreements with North Korea, Belarus and the de facto states of Abkhazia and South Ossetia. China also provided security guarantees to North Korea through the Treaty on Friendship, Cooperation, and Mutual Assistance. Recently, Saudi Arabia and Pakistan signed their own Strategic Mutual Defenсe Agreement.

Even the strongest security guarantees have limitations. For example Article 5 of NATO’s Washington Treaty says that an armed attack on one ally shall be treated as an armed attack on all, and commits each party to assist the state attacked with “such action as it deems necessary”. It contemplates that these actions may include “the use of armed force” but does not require an armed response. The treaty also recognises that member states may have internal requirements that need to be satisfied in order for force to be used, with Article 11 providing that its provisions will be carried out in accordance with the parties’ “respective constitutional processes”. Still, Article 5 is widely understood to be at the core of the Euro-Atlantic security architecture. 

By contrast, Article 42.7 of the Treaty on European Union – while an expression of mutual support – does not carry the same weight for its members. Article 42.7 of the treaty says EU member states shall respond to aggression to any bloc members “with aid and assistance by all means in their power”. But it also says that for NATO members (ie, all EU member states save Austria, Cyprus, Ireland and Malta), NATO is “the foundation of their collective defence and the forum for its implementation”. This reflects a certain reality about the bloc’s primary focus. While in recent years the EU has taken on greater security and defence roles, it is not fundamentally a military alliance like NATO.

It is possible to find treaties that are relatively explicit about wartime obligations, but that specificity may also be offset by conditional language. For example, Article 4 of the Treaty of Comprehensive Strategic Partnership between Russia and North Korea stipulates that if one party is attacked, the other will immediately provide military and other assistance by all means available and in accordance with their laws – a phrase that leaves considerable room for manoeuvre. Russia’s bilateral treaty with Belarus states that in the event of an attack on Belarus using nuclear weapons or other weapons of mass destruction, or in the event of a conventional attack that critically threatens Belarus’ sovereignty or territorial integrity, Russia may use nuclear weapons in response.

Whether a security guarantee is sufficient to dissuade a prospective adversary from an attack (and reassures allies that they will be defended), hinges not just on the promise of protection, but the capacity of the guarantor(s) to undermine an adversary’s attack or punish it, as well as the credibility of their threat to do so. In part to strengthen that commitment, the U.S. makes it easier for itself to respond and be involved from the start of any fight by deploying troops and weapons, including in some cases nuclear weapons, on the territory of its allies. Arguably, announcements Russia had deployed nuclear weapons in Belarus were meant to send a similar signal. Experts often argue that historical behaviour, such as participation in past wars, also bolsters credibility by demonstrating willingness to fight on behalf of allies. While guarantees are carefully worded to create flexibility on the part of guarantors, if a guarantor chooses not to respond forcefully in the event of an attack on an ally, it may cast doubt on its commitments to other states.

Since before Russia’s full-scale invasion began, Ukraine has sought a security guarantee from its trans-Atlantic partners in the form of NATO membership. Since 2019, accession to the alliance and to the European Union – with its own mutual guarantee – have been enshrined as goals in Ukraine’s Constitution. Kyiv believes that NATO membership would deter further Russian aggression and ensure long-term peace. Thus far, however, no country has extended binding guarantees to Ukraine.

One reason Ukraine specifically wants legally binding guarantees is its bitter memory of the non-binding security assurances it received from Russia, the United States, and the United Kingdom as part of the 1994 Budapest Memorandum. This document committed the other three signatories not to violate Ukraine’s territorial integrity, and contained a pledge that in the event Ukraine faced a threat, they would consult with each other to define a response. Russia violated this arrangement when it seized control of Crimea in 2014 and fomented war in Donbas that same year. It did so again when it undertook a full-scale invasion of Ukraine in 2022, and staked a claim to more of its territory months later. (In 2014, the U.S. and UK met their obligations with a consultation that Russia did not attend.) Whether or not Russia would have behaved differently had the Budapest document been legally binding is unknowable, but Ukraine took from the experience a sense that assurances are unreliable, and binding guarantees, which should promise more than consultations, are the coin of the realm. 

More than that, however, Ukraine specifically seeks the Article 5 protection enjoyed by NATO members. Much as it seeks EU membership for purposes of economic and political integration with its European neighbours, Kyiv does not see the Article 42.7 mutual defence commitment as adequate to its security needs. For the reasons noted above, Ukraine’s political leadership sees it as lacking the authority and capacity to enforce collective defence commitments or facilitate military action. 

Still, Kyiv appears to have taken on board that NATO membership is likely out of reach for the foreseeable future. It has therefore tried to identify other ways to ensure that its current supporters will offer help in the event of future attacks – and will do more than they have done in the past. Ukraine wants its backers to send a clear message to Moscow that they will provide comprehensive assistance and that Russia will pay a high price if it invades again. Although it has become more cautious about how it frames its asks since a February 2025 Oval Office meeting blew up over the issue of guarantees, Kyiv continues to press NATO members to develop a plan for providing them. What it prizes first and foremost is a commitment from the United States, which it wishes to have in its pocket before it engages in negotiations with Russia, so that the latter is required to accept it as part of any final settlement.

At the same time as it seeks guarantees, Kyiv also wants its Western partners to commit to long‑term funding of its armed forces, realising that the threat from Moscow will determine its security for years to come. Ukraine and its backers agree that absent NATO membership or similar guarantees from key NATO members, the best insurance for the country’s postwar security will be a well‑armed, sizable Ukrainian military with a range of deterrent capabilities, including long‑range systems. Ukraine wants its army to be integrated into joint defence planning with partner states and backed by a strong domestic defence industry. Ukrainian estimates put peacetime costs of maintaining that force (excluding weapons production) at €50 billion per year – a substantial figure even though less than its current wartime costs of some $120 billion per year. 

For Ukraine to join NATO, all the existing members of the alliance would have to agree. Various NATO members have had different positions on the prospect over the years but there has never been a consensus in favour. At the alliance’s 2008 summit in Bucharest, the allies pledged that Ukraine (and Georgia) would eventually become members, but did not offer a timeline. This position was the product of an effort to bridge the gap between a United States that then strongly supported Ukrainian accession, and the large number of European allies opposing it. Today, a number of NATO members, such as France and the UK, publicly favour NATO membership for Ukraine, albeit only once the war ends. Others, including the U.S., Hungary and Slovakia, as well as Germany, are unwilling to make that commitment. No one has officially come out in favour of Ukrainian accession while fighting continues. 

NATO members that are hesitant to offer membership worry that mutual‑defence commitments to Ukraine could pull them into a war with Russia, with the attendant escalation risks, including nuclear use. Wartime guarantees are especially fraught because as long as Ukraine is fighting off a Russian invasion, such guarantees would be commitments to join the fight. As for peacetime guarantees extended after the war’s end, proponents argue that these would deter Russia and thus would never have to be acted on. Sceptics, however, point out that the unwillingness to fight for Ukraine now may undermine the credibility of future commitments to do so. After all, Russia could conclude that promises aside, when faced again with the prospect of war, Ukraine’s backers would refrain from fighting, just as they are doing today. Should Russia test this assumption and NATO choose not to jump into the fight, it could undermine the credibility and deterrent effect of Article 5. 

Related problems arise for governments entertaining bilateral or small group commitments to Ukraine. Among the topics on the table for the 33 countries terming themselves a “coalition of the willing” to aid Kyiv are other types of security guarantees for Ukraine similar to NATO’s Article 5. But if such agreements result in one or more NATO member states at war with Russia, other members could face pressure to also join that war, creating the same problems discussed above. 

There is no consensus on this question, even among “coalition of the willing” members. While Finnish President Alexander Stubb argues that security guarantees by definition imply that guarantor countries should be ready to enter into military conflict with Russia, French President Emmanuel Macron holds that guarantees should instead restore strategic ambiguity so Moscow cannot predict how Ukraine’s supporters would respond to an attack. Macron and others also believe that U.S. involvement is critical for deterrence to work, since Moscow seeks to avoid direct confrontation with Washington above all. But U.S. policy on this matter is unchanged since 2014: it has ruled out any force presence and any direct military conflict with Russia over Ukraine.

Against this backdrop, Ukraine’s European backers have sought other ways to make pledges that are substantial, but short of mutual defence commitments. These include rapid consultation mechanisms; consistent long-term and emergency assistance, including arms and ammunition supplies; and the snapback of any loosened sanctions should Russia launch a new attack. The non-binding commitments developed by the G7 countries in 2023 could serve as a model but, of course, remain non-binding. 

In addition, 26 of the 33 “coalition of the willing” countries have developed a plan to deploy troops after fighting ends to territories in government-controlled Ukraine behind Ukrainian lines. The hope is that this “tripwire” deployment might deter future Russian attacks because breaking through the line would be an armed attack on NATO troops. The plan has two Achilles’ heels. One is that it relies on U.S. involvement, which Washington seems unlikely to offer. At minimum, coalition members are looking to Washington to contribute air defence, intelligence and logistical support. At most, the coalition seeks a commitment that the U.S. will provide air support if Russia attacks NATO allies. The second is that it is hard to imagine Moscow agreeing to such a deployment if fighting ends in any sort of negotiated settlement, and easy to imagine the Kremlin testing it if the force is deployed without a deal.

Wary of conflict with Russia, and the potential for such a war to escalate to nuclear use, the U.S. was reluctant to offer robust security guarantees to Ukraine long before the full-scale invasion and remains so. U.S. President Donald Trump – whose team apparently hopes over time to rehabilitate Moscow and cultivate stronger trade relations with it – says he is unwilling to discuss U.S. participation in such guarantees. In his most recent discussion with Zelenskyy on 17 October, Trump avoided taking a clear position on the topic. Moreover, according to some reports, Trump has said that security guarantees should be provided to both Ukraine and Russia, thus potentially aligning with the Kremlin’s position, as described below. 

The Trump administration’s position on other NATO members’ future military presence in Ukraine follows a similar arm’s length logic. The administration welcomes the coalition’s intention to deploy forces to Ukraine following a ceasefire – a proposal that seems to align with Washington’s stated desire that European countries take the lead on Ukraine’s security future. At the same time, it shows little or no interest in making the kinds of commitments that may well be necessary to the deployment’s viability. Indeed, when it comes to deployments, Washington appears to be floating ideas of its own, which do not, at least as described, involve security guarantees per se. 

The kind of security guarantees that Moscow wishes to see would entrench Russian dominance over the territories it occupies, institutionalise Ukraine’s effective demilitarsation, and prevent other states from coming to Kyiv’s aid in the event of a new Russian attack. As such, this view of security guarantees is very different from the collective security commitments discussion described above. 

Russia has long held that the only security guarantees for Ukraine should be those incorporated into a comprehensive peace agreement, alongside what it describes as security guarantees for Russia. To operationalise this, Moscow envisions that the five permanent members of the UN Security Council – including Russia – and any other regional states (Moscow has identified Belarus and Türkiye in the past) would pledge to respond to attacks on Ukraine under clearly delineated conditions and only if they all agree. 

Moscow and Kyiv discussed this approach during their failed negotiations in Istanbul in early 2022. Specifically, according to a draft agreement from that period, if Ukraine is attacked, these states would be required to hold consultations within three days and decide whether to provide assistance. Consensus would be required for any affirmative decision, meaning that Russia would be able to block any action. Moreover, the guarantors could use armed force only defensively and only on territory Ukraine controlled at the time the agreement was signed – when battle lines were close to Kyiv. 

Further, enforcement of these guarantees would be contingent on Ukrainian compliance with neutrality requirements also to be spelled out in binding provisions. These include requirements for Ukraine to remain non-aligned, without nuclear weapons, subject to strict numerical limits on the size of its military and, according to sources close to the negotiations, without long-range strike capabilities. Russian sources close to the negotiations also told Crisis Group that Russia would like Ukraine to commit to not using military force to liberate its occupied territories. Russia also wants guarantees in writing that NATO will not add members that lie east of its current membership.

Taken together, these limitations on Ukraine (and NATO) are what Moscow thinks of as its own security guarantees. By way of precedent, Moscow points to the 1955 Austrian State Treaty signed by the United States, the Soviet Union, the United Kingdom, France and Austria. That treaty granted Austria its post-WWII statehood in exchange for limits on its armed forces and military industry and its commitment to neutrality.

Russia may have another aim in its vision for security guarantees: enshrining its own decisive role in European security decision-making, a goal it has pursued since the early 1990s. After the Soviet Union’s collapse, Russian foreign policy centred on obtaining agreement from trans-Atlantic states on a European security mechanism for joint decision‑making – one that would guarantee a Russian voice in, and ideally veto over, the decisions other regional states made. Over the years, Moscow has tried and failed to attain this through the OSCE; cooperation with NATO and the EU; and with a proposal for a new treaty on European security. The Kremlin may now see a collective decision‑making model for Ukraine’s security, involving Russia, the U.S. and several key European countries as the template for broader decisions on European security in the future. Power dynamics would have to shift dramatically, however, for most European states to entertain this. 

The various visions for security guarantees and assurances, be they immediate or post-conflict, are clearly incompatible, and what happens on the battlefield may yet shift perspectives and change what is possible. At present, however, it is hard to understand how the discussion of security guarantees helps illuminate a path forward for Ukraine, especially given that the most plausible way forward is one in which Ukraine’s backers pledge continued aid, not fighting forces. In this scenario, the primary deterrent to future Russian aggression would be Ukraine’s own force strength. While Moscow will surely try to limit Ukraine’s military and its relationships with partners, it is unlikely that a Kyiv that has not been defeated will accede to any plans that undermine its long-term deterrent capacity. That requires a capable military underpinned by enduring partner commitments to armaments, training, intelligence cooperation, and defence financing. Such cooperation would demand institutionalisation of joint defence planning, deepening Ukraine’s integration into the Euro-Atlantic security framework short of NATO membership.

One way or another, getting to any deal other than Ukraine’s surrender requires Moscow to want an end to the fighting enough to be willing to temper its aims, at least temporarily. Formulas that pretend to guarantee Ukraine’s security, but in fact do not, are not the path to get there.