The European Court of Justice strikes against transparency again. Our response should be to call for an end to corporate limited liability 

Our beneficial ownership lead and acknowledged international expert Andres Knobel has had it. As another court decision supports the ‘weaponisation of privacy’ to defeat even basic transparency, Andres argues that the lobbyists have been so successful that they have defeated the case for the longstanding quid pro quo that justifies limited liability. Is it time for the return of unlimited liability? 

“Saudade não tem fim” (nostalgia has no end) could describe the feeling among transparency activists today. The momentum that started with the Panama Papers a decade ago and positioned Europe at the vanguard of beneficial ownership transparency is undoubtedly gone following the infamous European Court of Justice ruling of 2022 that invalidated public access to beneficial ownership information. But it did not end there. The weaponisation of privacy is now sweeping away transparency regulations that predate even the advent of beneficial ownership transparency. On 3 September 2026, the European Court of Justice managed to sink even lower by invalidating public access to shareholder information. 

This counter-secrecy reform based on the weaponisation of privacy is not just affecting beneficial ownership data or public access to information. Since 2022, rulings by the European Court of Justice and by the European Court of Human Rights have invalidated access by tax authorities to banking information as well as to company formation data held by law firms. But let’s get back to the issue of beneficial ownership and shareholder information. 

The ruling’s flaws in practice 

The September 2026 ruling that invalidated public access to shareholder information follows the same line of argument as the 2022 ruling that invalidated public access to beneficial ownership information: the general public does not need access to shareholder information because competent authorities and obliged entities (eg banks and lawyers) are the ones responsible for fighting money laundering and other illegal activities (para 79). For this reason, according to the Court of Justice, it should suffice for access to shareholder information to be based on demonstrating a legitimate interest, just as is now the case for access to beneficial ownership information in the EU (para 89). 

The ruling’s arguments may sound sensible in theory, but they do not apply to real life. First and foremost, “legitimate interest access” does not work in practice. “Legitimate interest” may sound like a good balance between privacy and the public interest, but it is usually a way to deliberately or inadvertently restrict access. For instance, in Argentina, where shareholder information had always been publicly accessible, the Commercial Registry decided to restrict access to shareholder information based on a legitimate interest when a former vice president came under investigation for corruption. Journalists and legislators investigating the vice president were found not to have a legitimate interest and had to go to court to request access. In the EU, the 2022 Court of Justice ruling reinstated access to beneficial ownership information based on a legitimate interest (as required in 2015 by the 4th Anti-Money Laundering Directive), even though legitimate interest access had already proven not to work in the past. It was because legitimate interest was ineffective that, in 2018, the EU established public access to beneficial ownership information, until the Court invalidated it in 2022. Unfortunately, the second time’s not the charm. Transparency International reported in 2025 that legitimate interest access to beneficial ownership information was not working properly in most EU countries. 

Second, it is disheartening that the Court did not consider the impact, or lack thereof, of its rulings in real life. Despite shareholder information having been publicly available online for decades in a number of countries, from Latvia and the UK to Ecuador and New Zealand, without any evidence of misuse, the Court made it clear that “it does not matter whether the information in question relating to private life is sensitive or whether the persons concerned have been inconvenienced in any way on account of that interference” (para 67). For the Court, the mere fact that the information is publicly available is bad enough. 

Nor does the Court consider whether authorities have the staff and resources to implement legitimate interest access and respond to requests for access to beneficial ownership information, and now also to shareholder information. The ruling makes clear that a theoretical infringement of the right to privacy outweighs any practical difficulties faced by authorities: “although the referring court indicates that the national companies register, which is the data controller, might not be in a position to determine whether each person requesting information does in fact have a legitimate interest in accessing the personal data concerned, it should be borne in mind that any practical difficulties associated with verifying the existence of a legitimate interest are not such as to demonstrate that an interference with the fundamental rights guaranteed by the Charter is strictly necessary” (para 83). 

What our arguments against the new ruling could look like 

As mentioned above, the Court’s arguments for invalidating public access to shareholder information are similar to those used to invalidate public access to beneficial ownership information. Our work on privacy washing and the weaponisation of privacy also applies here. 

1. On privacy and the right balance 

The first argument we could repeat is that “private family life” should stop, or at least be limited, when an individual goes outside their private family home and engages in “public” acts such as coming before a government authority to create a company that can sue other people, own assets with the protection of private property (and thus exclude others), limit the liability of its members against all of society and sometimes even get a bailout from the government. None of these corporate acts are “private” matters. Second, we could argue that information on shareholders would be unlikely to say anything about the wealth of those individuals (para 70), given that there is hardly any integration between asset ownership and corporate registries, and that ownership of an unlisted and unknown company says very little about one’s wealth. 

Thirdly, we could point to one contradiction in the ruling regarding public access to information on the initial shareholders of a company – the risk of pointing this out, however, is that even more transparency could be taken away in the future. The Court does not seem to have a problem with the fact that the initial shareholders of a company who signed the incorporation documents must be named and have their information made publicly available (Art 14(a) and 4(i) of Directive 2017/1132). It is only subsequent changes that need not be made public. In other words, the Court somehow considers that the privacy of new shareholders matters, but not that of initial shareholders – even if they remain shareholders because they have not sold their shares: “while Article 14(a) of Directive 2017/1132, which is to be read in conjunction with Article 4(i) of that directive, requires disclosure of the instrument of constitution and the statutes of a public limited liability company, those documents being required to state the identity of their signatories, including, where applicable, the company’s initial shareholders, no provision of that directive expressly requires disclosure of any subsequent change in the composition of the shareholders of such a company” (para 43). 

Finally, the Court claimed that “where there is a choice between several measures appropriate to meeting the legitimate objectives pursued, recourse must be had to the least onerous” (para 62). Based on the need to choose the less onerous measure when two rights are in dispute (eg privacy vs the public interest in information), the Court could have considered that countries are already implementing a “less onerous” measure that addresses the interference with privacy: corporate law allows shareholders to hold shares through other entities rather than directly in their own name. By holding shares indirectly, natural person shareholders would not be disclosing their information in public shareholder registers. 

2. On the importance of shareholder data for beneficial ownership transparency 

As regards the need to access shareholder information for beneficial ownership transparency, the following arguments could be proposed. First, shareholder information, including the full ownership chain of shareholders, is indispensable for verifying beneficial ownership information: if you don’t know who owns every layer of intermediate entities, you cannot confirm who the ultimate beneficial owner is. 

Second, shareholder information is especially relevant when there is no public access to beneficial ownership information, as is now the case in most of the EU following the 2022 ruling. In this case, journalists, civil society organisations and even authorities (especially foreign ones) need to determine the beneficial owner on their own, starting with shareholder information. Investigators need to check which natural person shareholders directly or indirectly hold more than 25 per cent of the shares. 

Third, although many beneficial ownership definitions consider anyone with more than 25 per cent of the shares to be a beneficial owner, access to shareholder information is necessary for investigations because the purpose of beneficial ownership is not just to determine who is in control of a company. As we have argued many times, a 1 per cent interest in a very profitable company may give the shareholder no control over the company, but it may be extremely relevant for tax and asset recovery purposes. That valuable shareholding without control may help determine a person’s wealth tax liability. It can also be used to determine unjustified enrichment if the shareholder cannot explain how they acquired that stake based on their low declared income, or it may help with the enforcement of debts if that minority shareholder owes money to authorities or other individuals. 

Lastly, data on all shareholders (including minority ones) is particularly relevant when investigating complex ownership structures. A minority shareholder may indeed be a beneficial owner, for instance in a scenario where two majority shareholders, each with 49.5 per cent of the shares, are in dispute. In this case, the shareholder with 1 per cent would have control because their vote would enable either of the other shareholders to obtain a majority. In addition, shares can have different classes, with one type giving extraordinary or exclusive voting rights. In this case, an individual could retain control over a company despite being a minority shareholder. More importantly, information on all shareholders is needed to uncover obfuscating strategies, for instance where there are more than four shareholders so that no one passes the 25 per cent threshold to be identified as a beneficial owner. (Interestingly, the plaintiff in the lawsuit that resulted in the 2026 ruling analysed in this blog post was a company with more than 17 shareholders.) In cases with many shareholders, beneficial ownership transparency requires all direct and indirect shareholdings to be considered to determine the beneficial owner. A minority shareholder can only be confirmed not to be a beneficial owner after confirming that they do not own additional shares either directly or indirectly. Companies that claim not to have a beneficial owner (eg because no individual has more than 25 per cent of the shares) should be thoroughly investigated because they could be deliberately distributing shareholdings among so many people that no one passes the beneficial ownership threshold. Instead of rewarding them with more secrecy, these companies with many shareholders should be considered high risk, and all of their shareholders should be identified and investigated. 

Yet it is unlikely that any of these arguments would suffice for a Court that has proven to uphold the weaponisation of privacy. The Court seems to have little regard for what happens in the real world, as it appears to expect that understaffed and under-resourced authorities (at best), or captured ones (at worst), will effectively fight financial crimes without the help, advocacy and demands for accountability from journalists and civil society organisations, which will no longer be able to easily access information. 

So it’s time to explore other arguments. 

What our arguments should be 

Lack of access to shareholder information will deal a heavy blow to beneficial ownership transparency because it will make it harder to determine and verify beneficial owners and to hold authorities accountable for a lack of enforcement. However, we should also acknowledge that most beneficial ownership registries and frameworks aren’t really working effectively, even in scenarios with full transparency, such as in the UK, where free online public access to beneficial ownership data has so far resisted the weaponisation of privacy. 

One ambitious solution could tackle both the weaponisation of privacy and low compliance with beneficial ownership transparency: ending corporate limited liability as we know it. 

There are multiple reasons for this poor state of beneficial ownership transparency. Many of these have to do with a lack of understanding, a lack of supervision, low sanctions and no enforcement in cases of non-compliance. But from a law and economics perspective, the main reason why beneficial ownership transparency isn’t working is the lack of incentives. Beneficial ownership is a transparency measure to help tackle illicit financial flows, but it is completely unrelated to corporate law. One need not be registered as a beneficial owner in order to control a company, receive dividends or vote. All of that depends mostly on shareholdings. For this reason, companies comply with beneficial ownership transparency only if they want to, or if they are afraid of sanctions for non-compliance. If companies don’t comply with beneficial ownership requirements, they can keep operating in the economy, enjoying limited liability and all the same rights as any other “subject of law”. 

Unlike beneficial ownership registries, which are now required to be strengthened and resourced to verify information (eg process discrepancy reports) and undertake many other proactive measures, commercial registries worked well even when they were passive repositories of information that accepted pretty much whatever companies declared. They worked, and continue to work, because companies have an incentive to tell the truth when it comes to declaring their directors and representatives who can represent the entity before third parties. It’s in their own interest. The same logic explains why asset registries work. People want to tell the truth to the land registry about who the legal owner of a house is, so that the registered legal owner can enjoy the protection of private property. (When legal ownership is different from beneficial ownership, the best of both worlds is achieved: enjoying the protection of private property without disclosing the real individual who owns and enjoys the property.) 

But, and this is a big but, public access to shareholder information was never relevant from a corporate law perspective (and that is why neither the Financial Action Task Force nor the EU Directive on Company Law requires it as basic information that should be public). Unlike “general partnerships” (eg société en nom collectif), where all partners are fully liable and thus their identity is important to third parties, in a typical company with limited liability (eg société anonyme), the identity of the shareholders is irrelevant (that’s why it can be called an “anonymous company”). Because of limited liability, the company will only respond with its own assets (not those of the shareholders), so whether shareholders are criminals or honest people, rich or poor, makes no difference from a liability perspective. 

The end of limited liability could thus be the answer to how the law could be changed so that incentives are aligned with transparency and compliance with beneficial ownership requirements. Our proposal is not just to require beneficial ownership registration as a precondition for any shareholder to vote or receive dividends, but to replace full limited liability with pro rata liability, so that shareholders and, where applicable, beneficial owners become liable for the entity’s obligations (based on the percentage of their interest). This way, the identity of shareholders would become relevant and would thus need to be publicly available. 

This proposal to replace limited liability with pro rata liability may create outrage among those who think this is throwing the baby out with the bathwater. However, the resulting transparency from pro rata liability is a very good side effect, not the main reason for it. The real reason to end limited liability is justice. 

The current status of limited liability is unfairly asymmetric: shareholders’ losses are capped at the amount invested in the firm, while their gains are unlimited. A revision of limited liability could propose that either losses remain fully capped (as they currently are), but gains are also capped, say at double the investment – with any gain beyond that deemed to belong to the State, similar to a windfall tax – or, more reasonably, that gains and losses are made symmetric. Currently, shareholders receive dividends on a pro rata basis. Liability should apply in the same way to shareholders, and if the shareholders are insolvent companies, then the beneficial owners would become liable. 

This is what we proposed in our paper Rethinking limited liability. In fact, the current state of affairs, where incorporation equals limited liability, is not a “divine right” set in stone, but rather a concept that emerged after the “separate personality” resulting from incorporation. First, there were many cases of firms with double or triple liability, or pro rata liability. Second, limited liability was conceived as a way to promote economic growth by encouraging individuals to invest in businesses without risking losing all of their personal wealth. Instead, limited liability is now enjoyed even by multinational corporations that create special purpose vehicles simply to isolate risks within the same economic group. 

While this proposal to replace limited liability with pro rata liability should be further explored to determine how it could be implemented without unwarranted consequences, it would be a positive change towards greater fairness (symmetry between rights and obligations, and between gains and losses) and transparency – the identity of shareholders and beneficial owners would become known to the public. 

Conclusion 

To sum up, the weaponisation of privacy keeps expanding and taking away even longstanding transparency measures that were available before the new transparency momentum started by the Panama Papers. This weaponisation of privacy looks set to continue apace unless new measures are tried. For this reason, we should explore more ambitious reforms, such as replacing corporate limited liability with pro rata liability. By making losses and gains fully symmetric, not only would we achieve greater fairness, but the identity of shareholders and beneficial owners would also become necessary information for third parties and would therefore need to be publicly accessible.