Tulip v van der Laan: The future of cryptocurrency in the United Kingdom

Introduction

In Tulip Trading v van der Laan¹, the Court of Appeal allowed an appeal against a jurisdiction challenge, finding that the argument that bitcoin developers owed fiduciary duties to users had a real chance of success. The Court did not decide that such a duty existed, only that there was a serious issue that required the facts to be established. Giving the leading judgment, Birss J recognised that Tulip Trading’s case boiled down to one central assertion: that “the decentralised governance of bitcoin really is a myth.”² In this article, we set out how this dispute arose, what may happen when the facts come to be established and how this will affect the future of cryptocurrency, and digital assets more widely, in the United Kingdom.

Tulip Trading is a company associated with Dr Craig Wright, who has initiated a series of legal actions around the world seeking to prove that he is Satoshi Nakamoto, the inventor of bitcoin. Dr Wright is a key architect of bitcoin satoshi vision (BSV), a cryptocurrency that is a hard fork of another cryptocurrency, bitcoin cash (BCH), which was itself a hard fork of bitcoin (BTC). These forks originated due to various doctrinal, philosophical and technological differences as to the correct development of bitcoin – without descending into the details (of these differences or what exactly is a hard fork, which we discuss below), the importance to the current litigation is that the proponents of BSV are in some sense in conflict with proponents of BTC bitcoin. The converse is not necessarily true: BSV is minuscule compared to BTC. At the time of writing, BSV is not completely valueless, but each BSV coin is worth about 0.1% of a BTC coin and trade flows of BSV have declined to negligible levels.

The Court of Appeal decision

Tulip Trading asserted that it owned a large amount of cryptocurrency contained in a wallet to which it no longer had the private key. It argued that the software underlying bitcoin³ should be amended so as to constitute it owner of that cryptocurrency. And it also argued that the developers of bitcoin owed it a duty to create and promulgate those amendments. The Court of Appeal held that an argument along these lines could be constructed: that the developers of a given network were a sufficiently well-defined group, that they had undertaken a role which involved making discretionary decisions and exercising powers for and on behalf of other people, in relation to property owned by those people and entrusted to the care of the fiduciaries, and that therefore the developers were fiduciaries with a duty to act in the interests of owners, even by introducing new code to the bitcoin software. Whether this argument would succeed would depend on the facts, to be decided later at trial.

In approving this argument, the Court of Appeal took a largely orthodox view of the definition of a fiduciary as someone who has undertaken to act for another in circumstances giving rise to a relationship of trust and confidence (see Bristol and West BS v Mothew⁴). The definition was applied objectively – the fiduciary did not have to see themselves as such. Although this definition was expanded incrementally, that did not preclude the Court from determining that a completely new situation, such as the current one, could involve a fiduciary relationship.

The Court of Appeal recognised that this argument would only succeed if Tulip established that bitcoin worked in a way that was surprisingly different to the orthodox view. And, in fact, Tulip only cleared the low hurdle of plausibility because the Court took a certain view of the basic facts of bitcoin. In particular, they started from the basis that bitcoin software was released via a single repository only amendable by certain developers with access to the password.

How bitcoin software is developed

Users of bitcoin access the network through a specific computer programme, the client software. The canonical version of this software is maintained in a widely used software repository. Miners also use compatible software that is similarly maintained in the repository. This software collectively determines what is recognised as a valid blockchain: the particular protocols that every block must fulfil, including arbitrary numbers included in specified parts of the block as well as the general rules about the size and composition of every block. 

The client software effectively allows the user to see the blockchain – to see every bitcoin holding – but what they see depends on what software they use. BTC bitcoin software will allow the user to see the BTC bitcoin blockchain and will reject all others as ill-formed. One might almost take the definition of any blockchain as what is seen using its client software. Similarly, miners using the corresponding software will add blocks to the blockchain that is seen by the client software.

Software development and the Court of Appeal decision

The starting point of the Court of Appeal was that the software used by clients and miners was in a specific place under the password-protected control of a small number of developers. This community might change as developers left and new ones were given access to the code, but the password requirement meant that it was to some extent a small, well-defined community. They took Tulip’s case as the assumption that a change to this code would be adopted by users and miners – they would have no choice if they wanted to participate in the bitcoin blockchain.  

Tulip’s argument is that users and miners are compelled by self-interest to adopt the new version of this software. The Court of Appeal assumed that a hard fork was the only response to this argument. But the relationship between this particular piece of software and the bitcoin ecosystem is more complex. There is nothing to stop any user or miner from staying with the old version or using a completely different version of the software. This does not necessarily lock the user out of the blockchain. That may be the case if the software introduces an incompatibility, but this does not always follow. Crucially, any other software that is compatible with the publicly-available protocol could be used by users or miners. They could even write their own proprietary software and they would still be mining and have access to exactly the same blockchain as everyone else, provided that their software used the correct protocols.

The software in the repository does not have any unique status. The protocols that underlie bitcoin are publicly-available and most of the software is open source⁵. That is, anyone can produce bitcoin software for users or miners without access to the password-protected repository, either for the existing BTC blockchain or to create a new blockchain or fork the existing blockchain. Anyone can fix a bug in the bitcoin software by modifying their own copy of the software in the repository and releasing it and users and miners could in many cases use this copy instead without disruption – only certain changes will create a hard fork. The position could be said to be more nuanced than the Court of Appeal characterise it, in reasoning that appears to reflect not only Tulip’s argument but also their own view: “without the relevant password (etc.) for the bitcoin software account … , no one else, such as a concerned bitcoin owner, could fix the bug. If a bitcoin owner identified a bug and wrote the code to fix it, that fix could still only be implemented if the developers agreed to do so in the exercise of their de facto power.”⁶

Numerous iterations of blockchain software exist and are used by different communities who thereby see their own blockchains. What determines their importance is not the origin or location of the software they use, but the size and influence of the community using a set of compatible protocols (embedded in one or more pieces of software).  

A split in an existing blockchain, a hard fork, that creates a new incompatible blockchain that forms an alternative descendant to the original blockchain, normally arises due to a split within the community. When BTC bitcoin forked to create BCH bitcoin and eventually BSV bitcoin, it was the community consensus that gave BTC its existing value. The existing ‘BTC’ ticker was kept by BTC bitcoin and BSV bitcoin had to adopt a new ‘BSV’ ticker – again, this reflects how it is perceived by users.

Without a substantial split in the community, any particular iteration of a piece of software does not lead to a hard fork in any meaningful sense. The community will simply use a version of the software, or an equivalent piece of software, that is acceptable to it. Soft forks occur all the time, where rival descendants of a blockchain are created and compete until one emerges victorious. This is part of the normal functioning of the community. Anybody can attempt a hard fork at any time by releasing their own software and, again, the normal functioning of the community absorbs these attempts without disruption.

Overall, the presence of password-protected software in a single repository does not necessarily get Tulip’s argument off the ground. There is more to bitcoin than a particular repository and hard forks. What Tulip need to show is that the miner and user community is a myth – in fact, there is no community consensus giving value to bitcoin, there are only a few all-powerful entities that determine what is the correct blockchain and use the corresponding software. In this analysis, the software itself is almost an irrelevance and the cryptographic security surrounding the blockchain an illusion: bitcoin is owned by whoever a few controlling corporations say own it. The Court of Appeal recognise this when talking about the alleged ‘myth’ of decentralisation, but their reasoning on password-protected software repositories does not support the conclusion that the argument that decentralisation is a ‘myth’ reaches the level of plausibility.

Bitcoin as property: A purely legal argument

The Court of Appeal did not address another legal argument that may be fatal to Tulip’s case. The fiduciary duty that Tulip argued for rested on bitcoin being considered as a form of property which the fiduciaries are responsible for looking after on behalf of the owners. The recognition of cryptocurrency as a form of property is now commonplace: see AA v Persons Unknown⁷ and a string of subsequent cases. This recognition rests on a characterisation of the properties of cryptocurrency that brings them within the definition of property. In particular, the cryptographic security and independence from any individual or group of individuals of cryptocurrency gives it a degree of permanence; this is also part of the reasoning behind the Law Commission’s proposal for the statutory recognition of cryptoassets as ‘data objects’ that exist independently of persons or legal systems⁸.

But Tulip’s claim contradicts the facts that justify classifying bitcoin as property. According to Tulip, the ownership of bitcoin is entirely within the control of a small class of software developers. Private keys, community consensus and cryptographic security are all irrelevant: if this small class of software developers code a change to the protocol that allocates a new owner to a specific bitcoin, this will occur as a matter of course. According to this analysis, bitcoin is not independent of any person and is not permanent – it more closely resembles a ledger kept by one person that may be altered by them arbitrarily. It would follow that bitcoin should no longer be classified as a form of property and the justification for imposing fiduciary duties on software developers in favour of its owners would also disappear. Essentially, the very facts that Tulip seeks to prove to show a fiduciary duty undermine the status of bitcoin as property, which is a precondition to establishing the fiduciary duty. It is unfortunate that this argument was not presented to the Court of Appeal, as it might have allowed it to reject Tulip’s case on purely legal grounds.

What happens if Tulip wins?

The real test for the central argument in this case may come after it is concluded. The ultimate measure of bitcoin decentralisation would come if Tulip were successful and the Court ordered a defined group of software developers to alter the blockchain protocol so that Tulip was recognised as the owner of a certain amount of bitcoin. The following scenarios might occur in relation to BTC bitcoin:

1. The software developers write code implementing the Court decision and upload it to the repository. The vast majority of the community ignores this change and continues with a different version of the software. 

This is a likely scenario, in that it allows the software developers to comply with the Court order while minimising the damage to BTC bitcoin. The different version would be the de facto BTC bitcoin and the Court-compliant version would very quickly have little or no value. Tulip would be the owner of the contested bitcoin in Court-compliant blockchain, but it would have little or no value. If the overwhelming bulk of the community adopts the different version and only a vestigial stub are left with the Court-compliant version, it is likely that this would not even be seen as a hard fork. Regular software updates that change the protocol are not necessarily seen this way, even if their technical effect is the same, unless they lead to a split in the community. 

It is difficult to see how any order of the Court could prevent this scenario from occurring, even if software developers implement the code change as ordered. There does not appear to be any basis to force the software developers to continue to work, as unpaid volunteers, on the Court-compliant blockchain. In the absence of any replacement, that version of bitcoin would fall into disuse. Conversely, there also does not appear to be any basis to prevent the software developers from working on any other blockchain, including the new BTC hard fork. And, even more clearly, there is no way to prevent miners or users from preferring the fork from the Court-compliant blockchain. In fact, miners and users would almost certainly be compelled to switch to the new fork or see the value of their cryptocurrency holdings evaporate (as to why, see version 3 below).

In this scenario, Tulip obtains legal title to the bitcoin, but its value is reduced almost to nothing.

This option would also be an embarrassment for those advocating for the use of English law as a robust commercial system suitable for international transactions. The effect of the Court’s order could be used to show that the reasoning of the Court was not sound.

2. The software developers refuse to implement the changes ordered by the Court.

There are two versions here. Either the software developers rely on avoiding effective enforcement by the Courts and continue to develop BTC bitcoin or they collectively stand down from their positions as developers of the repository identified by the Court. In both cases, Tulip does not become owner of the contested bitcoin. In the latter case, it would be futile for Tulip even to try to obtain the bitcoin by enforcing the Court order – their only remedy might be damages against individual software developers if they could construct a breach of duty claim based on their failure to continue in this voluntary role, which may not be straightforward.

Developers might not wish to stand down from the project and may prefer the first option. Bitcoin is inherently cross-border and pseudonymous. English law judgments are widely enforced by Courts around the world and the English courts are creative in enabling assets to be tracked down, frozen and used to satisfy their judgments. Nevertheless, enforcing a Court order to write a piece of code against a shifting group of largely anonymous developers dotted around the world – many of whom stem from a libertarian hacker background – may be a steep challenge.

In this scenario, Tulip does not obtain the bitcoin and is left with an expensive and complex task in trying to obtain them or compensation from the software developers.  

3. The software developers write code implementing the Court decision and upload it to the repository. The community adopt the new code and the change in ownership is reflected in the main BTC bitcoin blockchain.

This is ostensibly what Tulip are requesting, but it is unlikely to occur. If the change in code is accepted, it will demonstrate unequivocally that ownership of bitcoin is subject to the control of a tiny number of individuals. The value of bitcoin is based on its independence from interference by any individual and the robustness of the protocol secured by cryptographic means. If a majority of this consensus supported a change in protocol that contradicts these qualities, bitcoin would plummet in value – it would become worth no more than the word of the small group of people who determine its owners. Users and miners generally act in their own self-interest – that people tend to act in their own self-interest is the psychological truth that enables bitcoin to work – and adopting the new code would be the precise opposite. This is exactly why adoption of the new code by community consensus is extremely unlikely.

In this scenario, Tulip obtains the bitcoin but its value is reduced almost to nothing.

4. Tulip succeeds in showing that a fiduciary duty exists, but the Court decides that – even if Tulip is the owner of the contested bitcoin – that fiduciary duty does not oblige the developers to transfer the bitcoin to it.

At first sight, this might seem contradictory. But, in fact, given the effect of transferring the bitcoin in scenario 3, it is perfectly plausible that the correct choice of a fiduciary would be not to implement any protocol change as it would damage all bitcoin holders and not benefit Tulip, because its bitcoin would become worthless.

In this scenario, Tulip does not obtain the bitcoin. The theoretical mismatch between the duties imposed by the English courts and the way bitcoin operates would remain, ready to be triggered in a future case. This might be an attractive option for any future Court as it avoids any direct conflict while maintaining putative judicial oversight. The disadvantage is that it creates a new layer of legal uncertainty around the interaction between English law and cryptocurrency. 

Why is Tulip bringing this action?

One fact becomes clear when enumerating the possible ways in which Tulip might succeed: in none of them is it likely to obtain the valuable asset it is asking for. Even in those scenarios where its legal title is restored, that value then becomes worthless. This raises the question: why is Tulip pursuing this expensive and lengthy legal action if it is unlikely to recover anything of value, even if it is completely successful?

One recent technical development may shed light on this mystery: BSV bitcoin has created the ability to modify the protocol directly to alter ownership of specific bitcoin, expected to be used to reinstate owners who have lost their bitcoin due to fraud or theft. In other words, BSV has already created exactly the mechanism which other blockchains might also be obliged to create if Tulip is successful. If that happens, although Tulip may not recover any valuable assets, the Court decision might be a useful vehicle for promoting the system of governance of the BSV blockchain.

Please click here to view the abridged version of the article published in IFLR.