In March 2026, the EU Omnibus Directive entered into force, significantly narrowing the scope and reducing the obligations of both the Corporate Sustainability Reporting Directive (CSRD) and the Corporate Sustainability Due Diligence Directive (CS3D). The package drew sharp criticism from civil society organisations and legal scholars, who condemned its substance as a dismantling of corporate accountability and questioned the legality of the legislative technique deployed. For those who had followed years of hard negotiations over mandatory human rights and environmental due diligence, it marked a moment of reckoning.

Most strikingly, the Omnibus stripped out the obligation to adopt a climate transition plan altogether by deleting Article 22 of the CS3D. Transition plans were already on uncertain footing under the original CS3D: Article 29, which established a harmonised civil liability regime for companies, never extended to them in the first place, itself a sign of how contested the climate dimension already was. The Omnibus then removed both pillars at once: the climate transition planning obligation itself, and the harmonised civil liability regime, leaving in its place a constellation of national civil liability regimes across Member States. 

But is it really over?

This piece argues that the EU’s retreat does not mark the end of corporate climate liability – it invites us to look elsewhere. Turn away from Brussels, and a different picture emerges: national courts quietly assembling, case by case, a functional framework for holding parent companies accountable for their group-wide climate strategies. The results are promising but fragile: dependent on the resources of individual claimants and offering no guarantee of consistency across jurisdictions. Yet litigation has one thing current legislation lacks: the capacity to reach inside corporate structures and hold specific actors accountable for specific conduct.

The Omnibus and the Closing of the EU Door

The CS3D, as originally conceived, promised to require large companies to identify and address the human rights and environmental impacts of their operations and value chains. After years of difficult negotiations, what emerged was already a significantly weakened text. The Omnibus went further still. The EU declined to create enforceable civil liability for climate planning failures, and has now removed even the planning obligation. That said, the picture is not one of permanent closure. The Omnibus also modified the existing review clause, providing for an evaluation in mid-2030 and every five years thereafter, covering in particular the effectiveness of the enforcement mechanisms put in place at national level, including their protective effects on rightsholders.

This matters because of what is lost: uniformity. A harmonised regime would have set baseline standards applicable across Member States, created a level playing field for businesses, and enabled claimants anywhere in the EU to access the same rights. Without it, climate liability reverts to a mosaic of national legal systems. Fragmentation is not without its own dynamic: a bold decision in one jurisdiction can travel, shaping arguments and expectations elsewhere. But that is not a strategy, and it is no guarantee for claimants in jurisdictions where courts remain cautious.

The retreat is not only a policy failure; it may also be a legal one. On July 23, 2025, the International Court of Justice delivered its first advisory opinion on climate change, unanimously confirming that states bear binding obligations under customary international law to regulate private actors’ greenhouse gas emissions, and that failure to do so may constitute an internationally wrongful act. Legal scholars and civil society organisations have argued that the deletion of Article 22 of the CS3D appears irreconcilable with that standard: a state that removes the only mechanism requiring companies to adopt credible climate transition plans can hardly be said to be acting with the “stringent due diligence” the Court demands. The Omnibus, on this reading, does not merely close a door – it may open another, before international and domestic courts.

The Turn to National Courts

The European Court of Human Rights has itself underlined the centrality of national courts in this space. In Verein KlimaSeniorinnen Schweiz and Others v. Switzerland (2024), the Grand Chamber emphasised the key role domestic courts have played and will play in climate litigation, noting that it falls primarily to national authorities, including the courts, to ensure that Convention obligations are observed. That role is now, by default, more important than ever.

One of the most significant corporate climate cases in Europe remains Milieudefensie and Others v. Royal Dutch Shell plc. (Netherlands). At first instance, on 26 May 2021, the District Court of The Hague ordered Shell to reduce its group-wide CO2 emissions by 45% by 2030 compared to 2019 levels, finding that the company owed a duty of care, grounded in human rights and the Paris Agreement’s goals, to align its conduct with that reduction pathway. On 12 November 2024, the Court of Appeal overturned this specific reduction order, holding that the 45% figure could not be derived with sufficient certainty from the scientific consensus. Yet it confirmed the underlying principle: that a parent company can bear a duty of care in respect of its climate obligations at group level. Milieudefensie appealed, and on 22 May 2026 the Dutch Supreme Court heard the case. It is the first time a supreme court anywhere has considered a company’s civil law obligation to reduce emissions. The Hoge Raad has not yet ruled. Whatever its outcome, that principle has begun to travel.

Even without a mandatory EU due diligence framework, national courts are increasingly willing to pierce corporate structures to identify who actually controls a group’s climate direction, and companies’ own disclosures are proving their most dangerous evidence.

The Subsidiary Shield, Rejected

In Greenpeace and Others v. ENI, Greenpeace Italy, ReCommon, and twelve Italian citizens sued ENI, Italy’s largest energy company, along with its main shareholders – the Italian Ministry of Economy and Finance and Cassa Depositi e Prestiti, both public entities – before the Court of Rome in May 2023, alleging that ENI’s fossil fuel production and insufficient decarbonisation strategy were incompatible with the Paris Agreement and sought an order forcing the company to cut its emissions. ENI contested the jurisdiction of the Italian courts, arguing that the conduct complained of was attributable to legally distinct foreign subsidiaries operating outside Italy, not to the parent company itself.

The Italian Court of Cassation rejected this framing on 21 July 2025. It recognised that the claimants were not seeking to hold ENI’s subsidiaries liable for their individual emissions. Rather, they were targeting ENI as parent company for failing to adopt a group-wide industrial and commercial strategy capable of delivering the necessary reductions in CO₂ emissions. The Court held that the fact that most of the disputed emissions emanated from subsidiaries was immaterial: it was ENI alone, as parent company responsible for elaborating and approving the group’s overall strategy, that was said to be the source of the alleged harm. Whether the subsidiaries’ emissions could ultimately be attributed to ENI – given their separate legal personalities and commercial autonomy – was a question for the merits, not for jurisdiction.

The Court’s reasoning, though procedural, signals a clear openness to a functional reading of multinational corporate groups: what matters is not formal structure, but who actually defines the strategy.

Own Words, Own Risk

Belgian courts took that logic a step further. In Hugues Falys and Others v. TotalEnergies, a farmer and three NGOs brought a climate claim directly against the parent company before the Commercial Court of Tournai. TotalEnergies contested jurisdiction, arguing that responsibility lay with its subsidiaries. The Court was not persuaded. In its ruling of 18 March 2026, it found that TotalEnergies’ own submissions – particularly those concerning its energy transition strategy and its commitment to the Paris Agreement – in fact demonstrated the opposite of what it intended: that TotalEnergies shapes the policy of the TOTAL Group and exercises de facto control over it. The Court noted that TotalEnergies’ annual reports explicitly stated that it “drives” the Group’s strategy, and that a review of the documents produced confirmed effective control over all Group companies, including on climate matters. TotalEnergies’ contrary argument, the Court held, did not withstand scrutiny. It was clear from the company’s own writings that it defines the Group’s strategy, sets climate policy, and oversees its implementation. The Court nonetheless stayed proceedings pending the outcome of a parallel case in France, Notre Affaire à Tous and Others v. TotalEnergies, citing the risk of irreconcilable decisions. TotalEnergies has since appealed the jurisdiction ruling; the case will now be heard before the Court of Appeal of Mons.

Annual reports and sustainability disclosures, designed to project ambition to investors and the public, were here repurposed as evidence of the very control that grounds legal liability. The more a parent company publicly claims to steer its group’s climate strategy, the harder it becomes to argue in court that it bears no responsibility for that strategy’s adequacy. One might ask whether this logic will prompt companies to disclose more carefully, or less. But silence carries its own risks: investors expect climate ambition, reporting obligations constrain what can be omitted, and courts may yet learn to read reticence as its own form of evidence. Notably, the Omnibus left intact the CSRD requirement to disclose a transition plan where a company already has one, meaning that the disclosure architecture that generates this evidence survives the deletion of Article 22.

Vigilance in the Gap

The French case raises, in substance, the same underlying question: whether TotalEnergies can be required to align its activities with a 1.5°C trajectory in the absence of a specific binding norm. Brought by Notre Affaire à Tous, Sherpa, France Nature Environnement, and the City of Paris under France’s 2017 duty of vigilance law, it was the first time that legislation had been invoked in a climate context. After years of procedural battles, the case finally reached the merits in February 2026. On 25 June 2026, the Paris Judicial Court held that TotalEnergies’s duty of vigilance extends to Scope 3 emissions, which account for roughly 90% of the group’s footprint. The Court reasoned that the causal link between energy production and the combustion of its products is established, and that TotalEnergies is in a position to exercise influence over those emissions. Scope 3 emissions from its subsidiaries therefore fall within the risks that the parent company must identify in its vigilance plan.

The Court declined, however, to set a specific production-reduction target, holding that the 2017 law empowers the judge to require a company to elaborate adequate vigilance measures but does not allow the court to substitute itself for the company by prescribing precise and detailed measures. It ordered TotalEnergies to supplement its vigilance plan within six months, with a further hearing scheduled for January 2027 to assess its adequacy. The Court stayed its ruling on the complementary claim under Article 1252 of the Civil Code, which empowers the judge to prescribe reasonable measures to prevent or put an end to ecological damage, holding that it could not rule on that basis until it knew what measures TotalEnergies would actually adopt.

The ruling is significant beyond its immediate outcome. The Court refused to read the Omnibus rollback as a step back for climate liability, and in doing so, it aligned with the argument developed in this contribution. Where TotalEnergies had argued that the removal of Article 22 and the obligation to establish a climate transition plan weakened the normative framework, the Court was unmoved. It held that the structure of the CS3D had not been altered and that its objective, reconciling economic development with environmental protection, remained unchanged. The repeal of Article 22, it said, related only to the climate transition plan and had no impact on the due diligence obligations set out in Articles 7 to 16 of the Directive. The political cut, in other words, did not necessarily land where the companies had hoped.

A Functional Approach – and Its Limits

What links these cases is a reading of the corporate group as an economic reality rather than a legal fiction. This functional approach to parent company liability has significant implications. It suggests that even without a European directive mandating due diligence, parent companies that publicly assert strategic leadership over their groups may face legal exposure before national courts.

This is not nothing. But it is not enough.

The case-by-case construction of liability through litigation is inherently slow. Each case must be fought from scratch – jurisdiction, applicable law, the merits – before any finding of substantive liability is reached. Both Greenpeace v. ENI and Hugues Falys v. TotalEnergies currently remain at that threshold. The process is expensive, demanding sophisticated legal teams, expert evidence, and years of proceedings. It is accessible primarily to well-resourced NGOs and claimants with the stamina and means to pursue it.

There is also the question of uniformity. Courts across Member States may reach different conclusions on functionally similar facts. Without a European framework to harmonise standards, the development of corporate climate liability risks becoming geographically arbitrary – advanced in some jurisdictions, absent in others.

And yet courts are not entirely blind to this risk. The decision of the Commercial Court of Tournai to stay proceedings pending the French judgment is itself an acknowledgment that fragmentation is a problem.

But this is cold comfort. Courts can stay proceedings; they cannot harmonise doctrine. They can learn from each other; they cannot bind each other. The fragmentation that the EU legislature chose not to address does not disappear because judges are mindful of it – it simply becomes their problem to manage, imperfectly and incrementally. That the burden of coherence has fallen to individual courts, rather than to the institution best placed to provide it, is perhaps the most telling consequence of the Omnibus retreat.

Sitting with the Tension

The EU’s Omnibus decision was a setback for those who believed that corporate climate accountability required a legislative foundation at European level. The deletion of Article 22 leaves a gap that litigation cannot fully fill.

And yet, courts are filling parts of it. Some decisions already demonstrate that national judges are taking seriously the question of parent company responsibility for their group’s climate conduct – and that companies’ own public communications may be their most consequential evidence. What a company says in its annual report about driving its group’s strategy is, it turns out, something a court will read carefully.

The picture that emerges is one of uneven, cumulative progress: legally significant, but structurally insufficient. Whether the EU will eventually return to the question of mandatory corporate climate liability – under political conditions more favourable to the ambition – remains to be seen. For now, that work falls to national judges, one case at a time.