On 14 October 2026, CNN reported that Russian missiles and drones have spent recent months striking civilian facilities in Ukraine belonging to Western-owned businesses: a soft-drinks bottling plant outside Kyiv hit in early September, a fast-food chain’s logistics warehouse destroyed by an attack drone on 25 September, production sites damaged in the Mykolaiv region in late August. The head of the American Chamber of Commerce in Kyiv told CNN these were repeated direct hits with drones and missiles, and Ukraine’s president publicly described one strike as a signal addressed to Washington. The story is trending in the United States today, and the legal question it raises travels well beyond American boardrooms. French groups in retail, agrifood, building materials and logistics rank among the largest foreign investors and employers in Ukraine, and a drone does not check the shareholder register before it hits. When a French-owned factory burns in the Kyiv, Lviv or Mykolaiv region, three questions decide whether the loss is absorbed or recovered: does the war-risk wording of the insurance policy answer, does the bilateral investment treaty between France and Ukraine open a road to arbitration, and can an award actually be turned into money from Paris, where Russian assets sit frozen but largely out of reach? This guide answers each question from verified French sources, statute, treaty and published case law, quoted word for word and linked so that you or your adviser can check every step. It closes with what the October 2026 European debate over a loan to Ukraine backed by immobilised Russian assets could, and could not, change for a private investor. Nothing below asserts responsibility for any particular strike beyond what the cited press sources report; legal characterisations are stated as avenues to plead, because only a judge decides.
I. A French-Owned Plant Hit in Ukraine: What Insurance and the Investment Treaty Really Cover
A. Will French war-risk insurance pay for a drone strike on a Ukrainian factory?
Start with the policy, because French insurance law starts with an exclusion, not a promise. Article L121-8 of the Insurance Code provides: “L’assureur ne répond pas, sauf convention contraire, des pertes et dommages occasionnés soit par la guerre étrangère, soit par la guerre civile, soit par des émeutes ou par des mouvements populaires.” Three words carry the whole negotiation: “sauf convention contraire”. Unless the contract says otherwise, losses caused by foreign war, civil war, riots or popular movements are not covered. A French group whose Ukrainian subsidiary relied on a standard French fire or business-interruption policy, without an express war-risk extension, may therefore discover after the strike that the default rule excludes exactly what happened.
The same article allocates the burden of proof, and the allocation is unforgiving for the insured when the war is a foreign one. Where the war risks are not covered by the contract, the insured must show that the loss stems from something other than foreign war, while it is for the insurer to prove that the loss stems from civil war, riots or popular movements. In plain terms, for a strike on a plant in Ukraine, which French law would analyse as foreign war from the French insured’s standpoint, the policyholder must prove the loss comes from something other than that foreign war, a difficult evidentiary position when the sky was full of attack drones. By contrast, if the insurer invokes civil war, riots or popular movements, it is the insurer that must prove that characterisation. The practical consequence is immediate: coverage for a Ukrainian site should be negotiated as an express, priced war-risk extension, with its territorial scope, sub-limits, waiting periods and exclusions for nuclear, chemical or cyber vectors spelled out, rather than inferred from silence.
French case law adds a second layer that in-house counsel often miss. The courts have long held that the war exclusion must be combined with the mandatory public-policy rule now codified at article L121-2 of the Insurance Code, under which the insurer guarantees losses caused by persons for whom the insured is civilly liable, whatever the nature or gravity of their fault. In a widely cited ruling on a pharmacy burned during the November 2005 urban violence, the Cour de cassation recalled the mechanism in these terms: the contracts contained no derogation from “l’article L. 121-8 du code des assurances, qui dispose que, sauf convention contraire, l’assureur ne répond pas des pertes et dommages occasionnés soit par la guerre étrangère, soit par la guerre civile, soit par des émeutes ou par des mouvements populaires”, and the court then examined how that statutory exclusion interacts with the mandatory guarantee of article L121-2 (Civ. 2e, 22 November 2012, joint appeals nos. 11-19.524 and 11-19.807). The lesson for a group operating in a war zone is structural: own-damage cover and liability cover obey different exclusion logics, and a loss adjuster cannot use the war exclusion to sweep away guarantees the statute makes mandatory.
Before any arbitration or asset hunt begins, the evidentiary file built in the first days decides everything that follows. Notify every potentially applicable insurer within the contractual time limit, including political-risk and war-risk carriers, and calendar each policy’s separate deadline. Commission a bailiff’s report, preserve debris and impact photographs, keep satellite imagery, air-raid logs, employee witness statements and the Ukrainian emergency services’ records, and separate in the file what proves the fact of the strike from what characterises its legal nature. Do not concede in correspondence that the loss falls under any particular exclusion; characterisation is a legal debate for later, and a hasty admission to a loss adjuster can be difficult to walk back. Have coverage counsel review whether the extension covers missiles and drones expressly, whether business interruption is valued on gross margin or fixed costs, and whether the policy requires operations to resume within a set time. This file will serve twice: once before the insurer, once before an arbitral tribunal.
One final insurance-adjacent point belongs here because the treaty itself anticipates it. Article 7 of the France-Ukraine bilateral treaty provides that where a contracting party’s regulations organise a guarantee for investments made abroad, that guarantee may be granted case by case to investments made on the other party’s territory. French investors should therefore check, alongside private war-risk cover, whether a home-State investment guarantee was secured before the loss, since cumulated private private and public cover changes both the recovery strategy and the subrogation picture examined below.
B. What the France-Ukraine investment treaty promises an investor who loses assets to war
The bilateral instrument exists, it is in force, and it was written for exactly this kind of nightmare. The agreement between France and Ukraine on the encouragement and reciprocal protection of investments, signed in Kyiv on 3 May 1994, was published in France by Decree no. 97-1106 of 22 November 1997, published in JORF no. 277 of 29 November 1997, after authorisation by the statute of 16 November 1995. Its protections cover French nationals and companies holding investments on Ukrainian territory, and three of its clauses matter most when missiles fall on a factory.
First, the expropriation clause sets a hard standard with a priced remedy. It bars expropriation, nationalisation and any other measure with direct or indirect dispossessing effect against the other party’s investors, other than for a public purpose and provided the measures are neither discriminatory nor contrary to a specific commitment. Any dispossession must then be compensated: the treaty requires prompt and adequate compensation equal to the real value of the investments, assessed by reference to the economic situation prevailing before any threat of dispossession, fixed no later than the date of dispossession, effectively realisable, paid without delay, freely transferable, and bearing market-rate interest until payment. A targeted destruction of a foreign-owned plant, if it could be attributed to a State and characterised as a dispossessing measure, would be pleaded against this standard, with valuation anchored before the threat materialised, plus market-rate interest until payment. Attribution and characterisation would be fiercely disputed, and no outcome should be presented as acquired before a tribunal rules.
Second, the treaty contains a dedicated war-losses clause, often overlooked, which guarantees a minimum standard of treatment rather than full compensation. Investors of either party whose investments suffer losses from war, armed conflict, revolution, national emergency or revolt on the other party’s territory are entitled, from that party, to treatment no less favourable than that granted to its own nationals or companies or to those of the most favoured nation. For a French investor whose plant is destroyed in Ukraine, this clause means Ukraine itself would owe treatment no less favourable than that granted to its own nationals or to the most favoured nation’s investors for war losses. It is a relative standard, not a blank cheque, but in a reconstruction economy distributing compensation funds and aid, relative standards are where real money is allocated.
Third, and most decisively, the treaty offers a direct road to arbitration that bypasses the local courts. Article 8 of the France-Ukraine agreement sends investment disputes between one contracting party and a national or company of the other down an amicable track first. If the dispute cannot be settled within six months of being raised by either disputing party, either of them may submit it to arbitration before ICSID, the Washington-based centre created by the 1965 Convention on the settlement of investment disputes between States and nationals of other States. The six-month amicable period is therefore not a courtesy but a jurisdictional clock: the investor should notify the dispute in writing, with a quantified claim and supporting evidence, and let the clock run while preparing the request for arbitration, because a premature filing risks a competence objection and a late one wastes leverage.
That this treaty-arbitration route works from Paris is not theory. In a 7 December 2022 ruling that every practitioner in this field should read, the First Civil Chamber of the Cour de cassation decided the fate of an award rendered in Paris against the Russian Federation in favour of a Ukrainian bank for Crimean losses (Civ. 1re, 7 December 2022, no. 21-15.390, Oschadbank). The Paris Court of Appeal had annulled the award on competence grounds; the Cour de cassation quashed that annulment. Its reasoning restated two pillars of French arbitration law. First, on the intensity of review: “si le juge de l’annulation contrôle la décision du tribunal arbitral sur sa compétence, qu’il se soit déclaré compétent ou incompétent, en recherchant tous les éléments de droit ou de fait permettant d’apprécier la portée de la convention d’arbitrage, ce contrôle est exclusif de toute révision au fond de la sentence.” Second, on the nature of treaty consent: “En matière de protection des investissements transnationaux, le consentement de l’Etat à l’arbitrage procède de l’offre permanente d’arbitrage formulée dans un traité, adressée à une catégorie d’investisseurs que ce traité délimite pour le règlement des différends touchant aux investissements qu’il définit.” The Court held that neither the arbitration offer nor the treaty’s investment definition carried a temporal restriction, and that the treaty’s temporal provisions stated a substantive rule rather than a condition of consent, so that the appeal court needed only to verify that the dispute arose after the treaty’s entry into force. For a French investor relying on the 1994 France-Ukraine treaty, the message is direct: the standing offer is real, the competence review in Paris is searching but not a retrial on the merits, and temporal objections must be measured against the treaty’s actual wording, not assumed.
II. From Paper Rights to Money in France: Enforcing Against a Foreign State and the Frozen-Assets Debate
A. How to enforce an award against a foreign State from Paris without tripping over immunity
Winning an award is one procedure; turning it into money in France is a second one, governed by French enforcement law and the law of State immunity. The starting point is welcoming. Article 1514 of the Code of Civil Procedure provides: “Les sentences arbitrales sont reconnues ou exécutées en France si leur existence est établie par celui qui s’en prévaut et si cette reconnaissance ou cette exécution n’est pas manifestement contraire à l’ordre public international.” Recognition or enforcement requires proof of the award’s existence and a single negative condition, no manifest contrariety to international public policy. A French investor holding an ICSID or treaty-based award against the State responsible for the loss would therefore apply for exequatur before the Paris judicial court, producing the award and the arbitration agreement, and enforcement follows unless the narrow public-policy exception applies.
The avenues of challenge are closed by statute, which is precisely what makes Paris attractive. Article 1520 of the Code of Civil Procedure opens annulment only in five cases: “1° Le tribunal arbitral s’est déclaré à tort compétent ou incompétent ; ou 2° Le tribunal arbitral a été irrégulièrement constitué ; ou 3° Le tribunal arbitral a statué sans se conformer à la mission qui lui avait été confiée ; ou 4° Le principe de la contradiction n’a pas été respecté ; ou 5° La reconnaissance ou l’exécution de la sentence est contraire à l’ordre public international.” Article 1525 adds that the decision on recognition or exequatur of a foreign award can be appealed within one month, but the appeal court may refuse enforcement only on those same five grounds. The Oschadbank litigation illustrates the discipline of this framework: the Cour de cassation censured the appeal court for converting a temporal question into a competence defect, quashed the annulment, sent the parties back before a differently composed Paris court, ordered the foreign State to bear the costs, and awarded the bank 3,000 euros under article 700. Annulment judges control competence fully, but they do not retry the merits, and drafting the arbitration request to fit the treaty’s definitions is therefore the decisive battle.
Enforcement against State property then meets the second wall: immunity from execution. French statute, shaped by the Sapin II reform, draws a bright line. Article L111-1-2 of the Code of Civil Enforcement Procedures provides: “Des mesures conservatoires ou des mesures d’exécution forcée visant un bien appartenant à un Etat étranger ne peuvent être autorisées par le juge que si l’une des conditions suivantes est remplie : 1° L’Etat concerné a expressément consenti à l’application d’une telle mesure ; 2° L’Etat concerné a réservé ou affecté ce bien à la satisfaction de la demande qui fait l’objet de la procédure ; 3° Lorsqu’un jugement ou une sentence arbitrale a été rendu contre l’Etat concerné et que le bien en question est spécifiquement utilisé ou destiné à être utilisé par ledit Etat autrement qu’à des fins de service public non commerciales et entretient un lien avec l’entité contre laquelle la procédure a été intentée.” Diplomatic and consular property enjoys near-absolute protection: article L111-1-3 bars enforcement on “les biens, y compris les comptes bancaires, utilisés ou destinés à être utilisés dans l’exercice des fonctions de la mission diplomatique des Etats étrangers ou de leurs postes consulaires”, absent express and special waiver. In practice, the creditor needs the enforcement judge’s prior authorisation, must identify property used for commercial rather than sovereign purposes with a link to the debtor entity, and must steer clear of embassy accounts, military property, cultural heritage and tax claims, which the statute lists as presumptively sovereign.
The substantive hook for a damages claim exists in French civil law, but it must be pleaded with its jurisdictional limits stated openly. Article 1240 of the Civil Code states: “Tout fait quelconque de l’homme, qui cause à autrui un dommage, oblige celui par la faute duquel il est arrivé à le réparer.” Against a foreign State for wartime destruction, this general principle could support a claim only if the French court has jurisdiction and no immunity from jurisdiction bars the suit, questions that turn on the nature of the acts and the defendant’s status and that a court alone can resolve. The realistic path for most investors therefore runs through the treaty and arbitration rather than a direct tort suit: document the loss, trigger the six-month amicable clock, arbitrate, obtain exequatur in Paris, then enforce against identified commercial-use assets with judicial authorisation. Each step is slow and none is guaranteed, which is why the insurance file and the arbitration file should be built in parallel from day one.
B. Frozen Russian assets and the European reparations loan: what a bombed-out investor can realistically expect
Clients invariably ask the same question: with hundreds of billions in Russian sovereign assets frozen in Europe, why can a destroyed factory not simply be paid from that pile? The answer starts with a distinction French enforcement lawyers repeat weekly: freezing is not confiscating. The European Union’s restrictive measures rest on instruments including Council Regulation (EU) No 269/2014 on individual listings and Council Regulation (EU) No 833/2014 on sectoral measures concerning Russia, which immobilise funds and economic resources and prohibit making them available to listed persons, without transferring ownership. France relays these freezes domestically through ministerial action: article L562-2 of the Monetary and Financial Code empowers the economy and interior ministers to order, jointly and for renewable six-month periods, “le gel des fonds et ressources économiques” belonging to or controlled by designated persons. A frozen account in Paris is therefore sterilised, not awarded to anyone, and unfreezing it in favour of a private creditor would require a legal basis that does not currently exist in French law.
The October 2026 debate clients read about in the press concerns a workaround at the European level, not a distribution to victims. According to press and think-tank coverage this month, including an October 2026 Carnegie analysis and October press reporting, EU Member States have revived a plan for a so-called reparations loan to Ukraine, disbursed in tranches for urgent budgetary and defence-industrial needs, serviced by the future flow of extraordinary revenues from immobilised Russian assets, with a European Parliament briefing framing the same choice between a reparations loan, a budget revision or an alternative solution. The Carnegie analysis, October 2026 press coverage and the Parliament briefing all describe a mechanism whose beneficiary would be the Ukrainian State, not individual creditors, and whose legal construction, immunity implications and allocation of risk among Member States remain contested. A French investor should follow this debate closely, because any eventual fund for war damage would likely run through Ukrainian compensation channels where the treaty’s war-losses standard could matter, but should not present the loan as an existing source of payment. It is a project, financed against future revenue flows, whose legality would itself be litigated.
Where French procedure does offer a private lever is on the listings themselves, for those on the wrong side of a freeze. European sanctions designations and their French implementation can be challenged before the EU courts and, for domestic effects, before the French judge, and recent General Court annulments in Russian-sanctions cases show that listings fall when statements of reasons are thin or evidence stale. Readers facing a frozen account or a blocked transaction in France will find the method, deadlines and evidentiary burden detailed in our guide to challenging an EU listing or a frozen account from France. The symmetry is worth stating plainly: the same Paris that enforces awards against foreign States also polices the legality of freezes, in both directions under the rule of law.
The action plan for a French investor whose Ukrainian site has just been hit therefore runs in four parallel tracks. First, preserve and quantify: bailiff reports, imagery, accounts showing pre-threat value, and separate files for insurer, arbitral tribunal and any future compensation body. Second, activate insurance and guarantee cover immediately, within each contractual deadline, without conceding characterisation. Third, open the treaty’s six-month amicable phase with a documented, quantified notice, while preparing the ICSID request so the clock produces leverage rather than delay. Fourth, map enforcement targets early: commercial-use property linked to the debtor, outside diplomatic, military, cultural and fiscal sanctuaries, with French judicial authorisation as the gate. Treaty subrogation supports this sequencing: where a home-State guarantee has paid out, the guarantor steps into the investor’s rights, and the treaty expressly preserves the beneficiary’s ICSID route: guarantee payments do not affect the beneficiary’s right to resort to ICSID or to pursue proceedings already brought before it until their conclusion. Parallel tracks, one file, no concession made lightly: that is how a burned factory becomes a recoverable claim rather than a write-off.
Conclusion
The CNN reporting of 14 October 2026 suggests that Western-owned civilian plants in Ukraine face deliberate, repeated strikes, and French-owned sites face the same skies. French law does not leave their owners without answers, but each answer has a price and a procedure. Insurance answers only if war risk was expressly bought, because the default rule excludes foreign war and puts proof of another cause on the insured. The 1994 France-Ukraine treaty answers with a relative war-losses standard, a prompt-and-adequate compensation rule for dispossession, and a direct road to ICSID arbitration after six months of documented amicable effort, a road the Cour de cassation’s Oschadbank ruling keeps open and disciplined in Paris. Enforcement answers only against commercial-use State property with prior judicial authorisation, behind the shields of diplomatic and sovereign immunity. And the frozen-assets debate answers, for now, with a European loan project for Ukraine’s budget, not a claims fund for private victims. No official source has established, at the date of writing, responsibility for any particular strike described in the press, nor the outcome of any claim a French investor might bring; contracts signed before the loss will outweigh any general analysis, and only the courts seized will decide. Build the file before the smoke clears, run insurance and treaty tracks together, and treat every characterisation as something to prove, not something to assume.