22 September, 2026
|
24 minutes to read
On 17 September 2026, a Turkish diplomat told Agence France-Presse about Ankara’s new draft memorandum for the Black Sea. Türkiye is proposing two protected shipping lanes for civilian vessels: a western route hugging the Romanian and Bulgarian coasts, and an eastern route through the territorial waters of the states on the sea’s eastern and southern shores. “We want to use the territorial waters of the adjacent states,” he said, “including Türkiye — and Georgia, if they are willing to take part.” Earlier, in October 2025, the EU convened a ministerial meeting in Luxembourg on cross-regional security and connectivity, including security in the Black Sea region. Georgia’s foreign minister, Maka Botchorishvili, was not invited at all.
It is clear that the Black Sea is being reorganized by Türkiye, Romania and Bulgaria, Ukraine’s drones, and the European Union’s money, and it is being reorganized without Russia. Georgia has been offered a seat at that table because of where it sits, not because of what it can do. The question is whether Georgia wants the seat and what it should cost.
The Black Sea of September 2026 bears almost no resemblance to the Black Sea of February 2022. Then, it was a Russian-Turkish lake with a limited NATO coastline. Now, as Ukrainian analysts at the Carnegie Endowment for International Peace argue, Russian operational control of the maritime battle space has fallen from roughly 90% to around 25%. Russia’s Black Sea Fleet lost its flagship in 2022, its main base in 2023, and, after Ukraine’s 12 August strike on Novorossiysk, saw the frigates Admiral Essen and Admiral Makarov disabled — most of what remained of its surface combat power this summer. It is a fleet that hides in port and is struck there.
And yet Ukraine is losing the export war. This paradox defines the current moment and must be the starting point for any serious Black Sea policy. Sea denial is not sea control. Ukraine has made the sea unusable for the Russian navy, but it cannot make Odesa’s cranes and silos safe from missiles and Shaheds launched from land.
And yet Ukraine is losing the export war. This paradox defines the current moment and must be the starting point for any serious Black Sea policy. Sea denial is not sea control. Ukraine has made the sea unusable for the Russian navy, but it cannot make Odesa’s cranes and silos safe from missiles and Shaheds launched from land. Russia struck the greater Odesa ports more than 180 times in the first half of this year and 67 times in July alone. Ukraine’s agriculture ministry has cut its grain export forecast; its storage will be full by November; the Danube alternative is choked by drought.
The Black Sea is therefore no longer a Russian–Ukrainian theatre with bystanders. It has become a shared sea in which every littoral state’s ships, platforms and pipelines are vulnerable to drone attacks.
The war at sea has also spread to everyone else. Since late June, at least 25 Turkish-owned vessels have been attacked. Turkish companies operate around 40% of the ships calling at Ukrainian ports, which exposes Turkish crews to being hit, and some have already been injured. A drone detonated a kilometer from a Trans-Balkan pipeline compressor station in Bulgaria on 8 August; Romanian jets destroyed another drone near the Neptun Deep gas platform on 20 August; Kazakhstan’s oil exports through the CPC terminal at Novorossiysk were suspended for a week in July after four tankers were hit in four days. The Black Sea is therefore no longer a Russian–Ukrainian theatre with bystanders. It has become a shared sea in which every littoral state’s ships, platforms and pipelines are vulnerable to drone attacks.
Türkiye’s Foreign Minister Hakan Fidan called for a mutual moratorium on attacks on commercial shipping in August, but Russia dismissed it as a “half measure.” Türkiye then quietly began withholding Dardanelles transit permits for merchant vessels bound for Novorossiysk — a step outside the Montreux Convention, which, inter alia, governs warships, and one of the most consequential things any state has done in this war without announcing it. Now comes the September memorandum, with its two lanes and its conditional invitation to Georgia.
The Black Sea is not short of institutions. It lacks institutions that work, and the reasons they do not work fall into three recognizable patterns.
The first generation, built between the end of the Cold War and the annexation of Crimea, was inclusive by design. The Organization of the Black Sea Economic Cooperation (BSEC), founded in Istanbul in 1992, gathered eleven states, including Russia, around a consensus rule. BLACKSEAFOR, a Turkish initiative of 2001, put six navies — Russia’s among them — into joint exercises and search-and-rescue; it froze after the 2008 war in Georgia and died after 2014. Operation Black Sea Harmony, Türkiye’s own maritime security patrol launched in 2004, invited Russia two years later and faded into oblivion after 2015. The Black Sea Grain Initiative of 2022 was the last of this line: a UN- and Turkish-brokered arrangement that worked for a year and collapsed the day Russia chose to leave it. The lesson of the first generation is that every format that included Russia functioned only while Russia found it useful, and Russia’s usefulness test was, as Luke Coffey and Can Kasapoğlu observed, whether it kept NATO out.
The second generation was Western, and almost all of its formats were very soft (or toothless). The EU’s Black Sea Synergy of 2007 and the Eastern Partnership of 2009 were regional-cooperation and neighborhood policies with no real security dimension capable of countering live threats. GUAM, the Georgia–Ukraine–Azerbaijan–Moldova grouping, was created in 1997 as an alternative to Russian-led integration and has never had the resources or the members’ attention to be more than a communiqué. The Three Seas Initiative of 2016 connected the Adriatic, Baltic and Black Seas for EU members only, admitting Ukraine and Moldova as partners but structurally excluding Türkiye and Georgia — the two states whose coastlines matter most. The Bucharest Nine, founded in 2015, gave the eastern flank a voice inside NATO but spans two seas with different threat geometries. NATO itself, after Warsaw in 2016, settled for a “tailored forward presence” in Romania — a multinational brigade and more air policing — after Romania’s proposal for a standing allied Black Sea fleet was cold-shouldered by Bulgaria, whose prime minister said he wanted to see sailboats and tourists in the sea, not warships. And the United States, whose Congress introduced a Black Sea Security Act in 2023, settled for report language obliging the State Department to write a five-pillar strategy that its own supporters described as loosely defined and unfunded, and that has not survived the change of administration in any recognizable form. The lesson of the second generation is that formats built in Washington, Brussels, or other European capitals either had no ships or had the wrong members.
The third generation of formats is wartime, functional and narrow. Ukraine’s unilateral corridor of 2023 replaced the Grain Initiative and has moved more than two hundred million tons without asking anyone’s permission. The Türkiye–Romania–Bulgaria Mine Countermeasures Task Group (MCM Black Sea), established in January 2024, has run ten operations under a six-monthly rotating command and, in July, amended its mandate to cover critical infrastructure. NATO launched Eastern Sentry in September 2025 after drones crossed into Polish airspace, and Romania has been shooting drones down over the Danube delta and the gas fields since. These are the formats that actually do things.
What included Russia did not survive Russia’s absence. What was broad had no ships or teeth. What has ships and arms is small. Nothing on the list is both wide enough to cover the whole non-Russian littoral and hard enough to protect what is being built on it and counter real threats.
Line the three generations up and the shape of the problem is obvious. What included Russia did not survive Russia’s absence. What was broad had no ships or teeth. What has ships and arms is small. Nothing on the list is both wide enough to cover the whole non-Russian littoral and hard enough to protect what is being built on it and counter real threats. That is the gap, and it is why something more ambitious than another conference, another platform, or another mandate amendment is now required.
It is tempting and wrong to say that the EU has no strategy. The European Union adopted its first Strategic Approach to the Black Sea Region in May 2025. It published an implementation report in July 2026. It launched a Connectivity Agenda Platform in June that is expected to mobilize up to two billion euros through international financial institutions for the Trans-Caspian corridor. It has agreed with Romania and Bulgaria to build a Black Sea Maritime Security Hub at Constanța and Varna. So, indeed, Brussels has a strategy, a budget line and institutional readiness.
What the non-Russian Black Sea lacks is not a strategy but a security instrument that matches it.
What the non-Russian Black Sea lacks is not a strategy but a security instrument that matches it. Of all the frameworks above, exactly one has put ships to sea to face live threats — the trilateral, the smallest, and the one that involves no Brussels institution at all. The EU’s own strategy acknowledges this: its security flagship is a monitoring hub, and the document says it will “coordinate with the existing trilateral demining initiative.” The strategy knows it needs a security layer it does not itself provide.
Brussels also has money. It can theoretically pay for infrastructure that runs across a sea in which drones are being shot down over gas platforms. The Georgia–Romania electricity cable alone is assessed at 3.57 billion euros; it would run 1,100 kilometers, 700 of them at depths beyond two thousand meters, within 150 kilometers of Russian naval bases. This is a serious constraint and requires insurance, which is very hard to obtain. Nobody will underwrite that cable against wartime risk at a price the project is currently estimated to cost.
The supply case is also conditional. Georgia was a net electricity importer in 2025, and almost all its exports were concentrated in May–July, limiting the dependable surplus available from its existing hydro-dominated system. Yes, Azerbaijan has begun delivering substantial renewable projects, but its 2022 pledge to develop 8 GW of wind and solar by 2027, with 80% intended for export, remains far ahead of its 2025 wind-and-solar production of just 0.77 TWh. The cable’s prospective supply therefore depends on additional plants, grid reinforcement and export arrangements being delivered simultaneously.
The same is true, on a smaller scale, of port investment in Georgia and Bulgaria, of war-risk premiums on every hull entering Odesa, and of the Neptun Deep platform, which is supposed to make Romania the EU’s largest gas producer from next year.
Europe, therefore, has strategy and money. It has no instrument that makes the assets it is funding defensible.
Europe, therefore, has strategy and money. It has no instrument that makes the assets it is funding defensible, and no political framework that binds the five non-Russian littoral states to defend them together. That instrument could be the Compact.
Vagueness is where regional-security proposals go to die, so when one is advocating for a Black Sea Compact, specifics matter.
The compact should be an agreement among the five non-Russian littoral states — Türkiye, Romania, Bulgaria, Ukraine and Georgia — with the European Union as a partner, to do four things together that none of them can do alone.
1) Protect critical infrastructure. Cables, pipelines, offshore platforms and port approaches, through shared surveillance, seabed monitoring and, where a member consents, patrol in its territorial waters. This is not a new idea; it is what the three-state Mine Countermeasures group agreed to do in July 2026 when it amended its mandate from mine clearance to “surveillance and reconnaissance… to prevent deliberate attacks against critical infrastructure.” The compact extends that mandate to five states and gives it a budget. The difference would be Ukraine’s participation, which could bring new technologies, capabilities, and a serious anti-Russian dimension to this endeavor.
2) Run protected lanes. Civilian shipping corridors along members’ territorial waters — the western lane along Romania and Bulgaria that already exists in practice, and the eastern lane along Türkiye and Georgia that Ankara has just proposed. The Black Sea Grain Initiative of 2022 showed that the model works; it also showed that a model which depends on Russian consent dies when Russia withdraws consent. Lanes inside members’ own waters do not. This, especially if Ukraine and Georgia are on board, would essentially leave Russia out of the Black Sea transit discussions. This would, however, require new arrangements on the illegal trade with Georgia’s occupied regions, which currently mainly happens from Türkiye. Several ideas on this were discussed about fifteen years ago, and bringing them into the Compact’s mandate would reduce Russian economic influence in the occupied regions.
3) Share maritime domain awareness. This would effectively mean the Constanța–Varna Hub, fused with Turkish coastal radar and Ukrainian battlefield intelligence. Türkiye supplies up to 67% of situational-awareness information to NATO and Ukraine in the Black Sea, while Ukraine brings extensive combat experience with maritime drones, electronic interference, and threats to commercial navigation. The hub should turn these complementary inputs into a continuously updated picture: identifying ships that disappear from tracking systems, checking suspicious movements near cables and gas platforms, and circulating verified warnings about drifting mines and unmanned craft. Permanent Turkish and Ukrainian liaison officers, a round-the-clock watch, and agreed rules for releasing sensitive information would connect detection to action, with each alert assigned to the national authority responsible for investigating or responding. Existing cooperation provides a foundation. The European Maritime Safety Agency has already demonstrated integrated maritime monitoring off Constanța, and Ukraine contributes combat lessons through NATO exercises. The next step is to make those exchanges routine and fast enough to protect shipping and infrastructure.
4) Underwrite the risk. On 16 September, the London insurance market’s Joint War Committee extended its listed area to almost the whole Black Sea, carving out only the territorial waters of Türkiye, Bulgaria, Romania and Georgia — so that every hull entering the open sea now needs separate war-risk cover at an additional premium, while every hull that stays inside those four countries’ waters does not. Basically, the underwriters have priced the compact before anyone has signed it: the littoral lanes are the only part of the sea the market still treats as insurable, and the eastern lane through Georgian waters exists on the JWC map before it exists in any agreement. Building on that pricing, an EU-backed war-risk reinsurance facility should give private insurers a defined public guarantee against losses on Black Sea shipping and infrastructure, with separate terms for individual voyages and for long-lived assets such as ports, cables and offshore energy installations. Existing mechanisms offer a starting point. The European Bank for Reconstruction and Development’s EUR 110 million guarantee facility supports war-risk insurance in Ukraine, and the Unity facility covers ships carrying non-military cargo to and from Ukrainian ports, but both are Ukraine-specific and voyage-based, and neither touches a cable, a platform or a Georgian port. A broader facility would need explicit coverage limits, deductibles, claims deadlines and cancellation provisions, so that lenders and operators can assess what happens after an attack rather than after a lawsuit. Alongside it, Brussels should broker a single corridor-governance agreement for the Middle Corridor’s Black Sea leg: interoperable electronic customs documents submitted once and reused by every participating authority; published through-rates covering rail, port handling and the sea crossing; coordinated schedules; and a designated operator responsible for resolving delays and disputes across the journey — the fragmented coordination and border procedures that the World Bank identified in 2024 as the corridor’s binding constraints. Brussels cannot deliver any of this alone, but it can tie guarantees and infrastructure funding to common service standards and transparent charges, and that gives investors and shippers what they currently lack in the Black Sea: an identifiable price, an accountable counterpart and an agreed allocation of risk — the three things the JWC has just told them they cannot get anywhere outside four countries’ territorial waters.
At the same time, the Compact cannot be everything; for instance, it cannot be a collective defense treaty; nobody should promise to go to war for anybody; however, its members could defend each other strictly at sea. It should not be a substitute for NATO or a rival to it. It should not deploy non-littoral warships, and it should not touch, revise, reinterpret, or reopen the Montreux Convention, even if many believe that it is now obsolete. The Compact should operate inside the sea, among the states that own its shores, on the principle of regional ownership. It should become, if you will, the trilateral mine-countermeasures group grown to five members, a wider mandate, European money, and a political charter. This would make it very similar to the security architecture that currently exists over the Baltic Sea.
Regional formats usually fail because they are talking shops. The remedy is a format in which each member’s contribution is specific, and the others need it.
Türkiye brings the only real navy in the basin, control of the Straits, the largest coastline, the strongest maritime industry, and the diplomatic standing to talk to Moscow and Kyiv at once. Without Türkiye, the compact cannot exist. In fact, Türkiye has every reason to spearhead such a Compact. It will reassert its leading role in Black Sea security, backed by European finance, without reopening Montreux or sharing authority over the Straits. Coastal states would provide the ships and conduct operations; the EU would help finance infrastructure, insure commercial risk, fund the Maritime Security Hub, and simplify corridor trade. This closely follows the division of labor proposed by Mustafa Aydın and Aslı Aydıntaşbaş, who argued for practical partnerships among coastal allies, coordination that respects Montreux, and European engagement through connectivity and investment.
Romania brings Constanța, the largest port on the sea and the EU’s Black Sea gateway; the Constanța half of the Maritime Security Hub; the expanding NATO base at Mihail Kogălniceanu, set to become the alliance’s largest in Europe; and, from 2027, Neptun Deep — an offshore asset large enough to make protecting it a European interest rather than a Romanian one. Romania also brings political entrepreneurship: the Security Hub was a Romanian proposal, pushed by Foreign Minister Oana Toiu from October 2025 until the Commission endorsed it in January.
Bulgaria brings continuity along the western shore and the capacity to make the compact less dependent on any single port or country. Its coastline lies between Romania and Türkiye: surveillance, mine clearance and incident response in Bulgarian waters are therefore part of the same coastal security problem. Sofia has already demonstrated a practical contribution, assuming command of MCM Black Sea in July 2026 and agreeing to acquire seven additional mine-countermeasure vessels from Belgium and the Netherlands. Varna’s half of the Hub could connect that expertise to shared surveillance, hydrographic information and maritime training. Commercially, Varna and Burgas offer additional gateways for grain, containers, industrial cargo and energy supplies, while Burgas’s refinery and terminals create their own protection requirements. Their value is not unlimited spare capacity: the EIB found that insufficient infrastructure prevented Varna from absorbing cargo when Constanța became congested, and backed a EUR 50 million grain-terminal investment to address that weakness. Bulgaria also connects the sea to alternative overland routes. The December 2025 Bulgaria–Greece–Romania corridor agreement envisages an eastern axis from Alexandroupolis through Burgas and Varna to Constanța, offering a basis for commercial diversification and military mobility. The compact should give Sofia specific, funded responsibilities, such as mine clearance, port resilience, repair support, and interoperable border procedures, whose delivery can survive changes of government. Bulgaria’s importance is the continuity of these services across the western coast.
Ukraine’s participation can be a game-changer. It is Ukraine’s naval drones, not any NATO deployment, that removed the Russian fleet from the equation. Ukraine brings operational expertise in a kind of warfare that no member navy has practiced; it brings the western corridor, which has moved more than two hundred million tons since 2023; and it brings, in the Sea Baby and Magura programs, an industrial base for exactly the unmanned systems that infrastructure protection now requires. Ukrainian defense analysts have argued that their country’s experience should be written into allied doctrine rather than admired from a distance. The compact could make that happen.
And Georgia? Georgia brings the eastern shore. That is the whole of it — and, as we will argue below, it is nonetheless enough. At least for now.
Georgia is not a Black Sea security provider. It has no navy, no mine-countermeasures capability, no integrated air defense, and no maritime surveillance beyond a coast guard. Russia occupies much of its coastline, and it is building a naval base at Ochamchire on that stretch. Georgia has not fought at sea, has not contributed to the trilateral, and — for two years — has not been invited to the Black Sea-related initiatives. Whatever Georgia brings to a Black Sea Compact, it is not force.
It brings territory, and three facts make that territory indispensable right now.
First, the Georgia–Romania electricity cable, the EU’s flagship connectivity project in the sea, lands at Anaklia. There is no Georgian-free version of it; the whole point is to connect Caspian and Caucasian power to the European grid through the Black Sea, and Georgia is the shore.
Second, the Middle Corridor has one Black Sea entry with rail behind it, and it is Georgia — Poti, Batumi and, if it is ever built, Anaklia. Kazakhstan’s president said in June that corridor volumes had risen fivefold to over four million tons, with a ten-million-ton target. Every one of those tons bound for the Black Sea must cross Georgia.
Third, and most immediately, Türkiye’s eastern lane runs through Georgian territorial waters. The alternative is a route that hugs the Turkish coast alone and stops short of the corridor’s ports — a route to nowhere.
But what if Georgia refuses? In fact, Georgia has refused to enter any arrangement that goes against Russia’s security interests. Does the whole edifice fall in such a case?
The answer is – no. The reason is that the Compact’s security core is on the western and southern shores, where the war is, where the ships are, where the MCM Black Sea already operates. If Georgia stays out, the western lane runs, the Hub opens, the trilateral becomes a quadrilateral with Ukraine, and Neptun Deep is defended. What fails is the eastern connectivity layer: the cable, the corridor’s Black Sea leg, the eastern lane. And the party that loses most from that failure is not Brussels or Ankara, but Georgia.
In fact, Georgia’s longer-term indispensability is real but is currently wasted by the Georgian Dream, which prefers to blackmail the West, arguing that “you need us more than we need you.” Friedrich Conradi argued at Carnegie in April that the Middle Corridor may prove “a dead end” because of Georgia’s political paralysis. If the West, and in this case the EU, plays its cards correctly, it could get Georgia on board and even unwind some undemocratic trends of recent years.
If the compact is vital and Georgia is needed, the obvious course would seem to be to drop the politics. The European Union does not condition its energy partnership with Azerbaijan on the state of Azerbaijani democracy, nor its corridor investments in Kazakhstan on Kazakh elections. It deals with both transactionally, and both sides understand the terms. Many ask: why should Georgia be treated differently — because it holds candidate status, or because it is small enough to be pushed around?
There are three reasons. The first is about what Georgians chose. Azerbaijan and Kazakhstan never asked to join the European Union. Georgia did. It negotiated an Association Agreement and a free trade area; it obtained visa-free travel based on specific rule-of-law commitments; it wrote European integration into Article 78 of its constitution; and in December 2023 it asked for and received candidate status, a relationship defined by conditionality. The conditions attached to Georgia are not a moral tax on a weak state. They are the content of the relationship Georgia itself demanded. Brussels does not apply them to Baku because Baku never signed up to them. To drop them for Georgia would not be treating Georgia like Azerbaijan; it would be treating Georgia’s own constitutional choice as a fiction.
The second is about Europe’s own interest. Deepening transit and infrastructure cooperation with Georgia at the price of dropping democratic demands might bring the European Union a short-term geoeconomic benefit. But Georgia’s drift towards the EU’s competitors, and their growing influence over the country’s strategic infrastructure, cast doubt on the long-term value of that benefit. In such conditions, European investment risks strengthening the position of the EU’s own rivals in the region. Upholding democratic principles and institutional requirements in relations with Georgia is therefore not a distraction from the EU’s economic and strategic interests. It is part of defending them.
A European Union that pays for a cable landing at Anaklia without conditions is not buying influence in Georgia, but buying a power line for a Chinese port.
Look at the evidence. The government that has spent two years dismantling the institutions Brussels asked it to build is the same government that handed the Anaklia concession to a consortium led by a U.S.-sanctioned Chinese state company, that upgraded its “strategic partnership” with Beijing this spring, and that has restored a working relationship with Moscow while shouting at Brussels. Democratic backsliding and the capture of strategic infrastructure by the EU’s competitors are not two problems. They are one problem seen from two sides. A European Union that pays for a cable landing at Anaklia without conditions is not buying influence in Georgia, but buying a power line for a Chinese port.
Conditionality on connectivity should be chosen because, for the first time since 2022, Brussels holds something Tbilisi cannot get elsewhere.
The third reason is that Georgia, under the Georgian Dream government, is not, in fact, easily pushed around — and the record proves it. This regime has ignored every European condition for two years, at the cost of frozen candidate status, visa-free travel suspended for its officials in March, the whole population warned in December that theirs is next, and its election unrecognized by the European Parliament by 444 votes to 72 in 2024. None of it moved Bidzina Ivanishvili. What is different about connectivity is that it is the one European good the Georgian Dream government actually wants — the ports, the cable, the corridor money, the investment that keeps the economy afloat and the patronage flowing. Conditionality on connectivity should be chosen because, for the first time since 2022, Brussels holds something Tbilisi cannot get elsewhere.
As for the conditions, they are well known, and we have covered them at length in this journal. In a nutshell, they include releasing political prisoners, refusing to ban political parties, rescinding non-democratic legislation, and reforming the electoral framework to allow disenfranchised citizens of Georgia to choose their governments in a normal competitive process.
None of these require a monitoring mission or a decade. Each is a decision the Georgian Dream could take next Monday, and Brussels could verify by Friday. That is what makes them appropriate as compact conditions and distinguishes them from the accession benchmarks. It is also what makes them honest: a government that refuses to take four reversible decisions cannot claim it is being asked for too much. It is telling everyone what kind of governance model it prefers for the Black Sea.
