18 April 2020

Cryptocurrences and Information as Property

Law unsurprisingly continues to catch up with 'information as property and the supposedly revolutionary concept of 'virtual currencies', ie private money systems.

In New Zealand the High Court in Ruscoe and Moore v Cryptopia Limited (In Liquidation)  [2020] NZHC 728 has concluded that cryptocurrencies are property.

Gendall J states
For present purposes it will become apparent that I reach the conclusion that the cryptocurrencies here situated in Cryptopia’s exchange are a species of intangible personal property and clearly an identifiable thing of value. Without question they are capable of being the subject matter of a trust. I will now set out my reasons for this conclusion. 
The authorities 
[70] This first issue outlined at para [46](a)] of this judgment asked the specific question whether any or all of the digital assets held by the liquidators are “property” within the definition outlined in s 2 of the Companies Act. 
[71] That section defines “property” as: ...property of every kind whether tangible or intangible, real or personal, corporeal or incorporeal, and includes rights, interests, and claims of every kind in relation to property however they arise. 
[72] Although there is a certain circularity with this definition, it is nevertheless inclusive and wide in that it extends “property” for the purposes of the Act to include “rights, interests, and claims of every kind in relation to property however they arise.” 
[73] Courts in New Zealand have accepted that the definition of “property” in the Companies Act is a “wide” one and includes “money” despite money not being expressly included in the terms of the s 2 definition. This is clear from the Supreme Court decision in McIntosh v Fisk. There, the Court accepted it was arguable that “the payment of money by RAM would fall within s 292(3)(a) as a transfer of property by RAM due to the wide definition of “property” in s 2 of the Companies Act.” 
[74] Further, in Chapman v Effective Fencing Ltd, Associate Judge Faire held:  
The definition of “property” in s 2 in referring to “every kind” of property, is wide enough to cover money.
Clearly money is “tangible” and “personal” property in terms of the definition. 
[75] Lord Wilberforce’s opinion in the House of Lords in National Provincial Bank Ltd v Ainsworth is often cited as the classic statement of the characteristics of “property”.  There, his Lordship said:
Before a right or an interest can be admitted into the category of property, or of a right affecting property, it must be definable, identifiable by third parties, capable in its nature of assumption by third parties, and have some degree of permanence or stability.
I will return to this definition shortly. 
[76] But first, I turn to several recent cases where the question of cryptocurrencies as “property” has been addressed to some extent. The first is a Singaporean case, B2C2 Ltd v Quoine Pte Ltd, which Ms Cooper in particular considered to be an important decision, given its factual setting was not dissimilar to the present case. 
[77] Initially this involved a 2019 first instance decision of the Singapore International Commercial Court, a new division of the High Court of Singapore created in 2015. That decision was then appealed to the Court of Appeal of Singapore and was the subject of a lengthy appeal judgment delivered this year.  In the lower  court, all parties accepted that cryptocurrencies were a species of “property”, a concession which the judge, Thorley IJ  accepted was rightly made. 
[78] The case concerned a Singaporean cryptocurrency exchange operated by Quoine, in many ways like Cryptopia, on which B2C2 was a trader. Some trading was set up to occur automatically through computers connected to the exchange and was pre-programmed. The transactions which led to the litigation were conducted by way of algorithms created by Quoine and by B2C2. The trades in question resulted from pre-programmed requests to exchange cryptocoins of ethereum for bitcoin. Errors occurred in the programming and an unusual set of circumstances resulted in B2C2’s computer offering ethereum for bitcoin at the rate of one ethereum for 10 bitcoin. The computer of another trader on that platform accepted that bid, seven such trades taking place (“the disputed trades”). The going rate of ethereum for bitcoin in the market at the time was one ethereum for 0.04 of a bitcoin. The effect of the automatic trading was that B2C2 sold ethereum at about 250 times its appropriate price. Quoine became aware of the mistake. It then reversed the trades which led to the litigation. 
[79] B2C2 sued Quoine in the High Court for breach of the contract between it as a trader and Quoine as the operator of the exchange and for breach of trust as a result of Quoine’s having returned the bitcoin to the counterparty. A defence of mistake was raised in that Court but Thorley IJ held there was no basis for setting aside the trading and Quoine was accordingly liable to B2C2 for having wrongly reversed the trades. He upheld both B2C2’s contract claim and its claim for breach of trust. 
[80] That breach of trust claim could have succeeded only if the bitcoins in question were an asset that could form the subject matter of a trust. At the lower court level, Quoine had conceded that Bitcoin was a species of “property” but it did not concede that there was any trust. Thorley IJ considered that the concession on the “property” point was rightly made and in his judgment his Honour stated:
Cryptocurrencies are not legal tender in the sense of being a regulated currency issued by government but do have the fundamental characteristic of intangible property as being an identifiable thing of value. Quoine drew my  attention to the classic definition of a property right in the House of Lords decision of National Provincial Bank v Ainsworth [1965] UKHL 1; [1965] 1 AC 1175 (HL) at 1248: "...it must be definable, identifiable by third parties, capable in its nature of assumption by third parties, and have some degree of permanence or stability". Cryptocurrencies meet all these requirements. Whilst there may be some academic debate as to the precise nature of the property right, in the light of the fact that Quoine does not seek to dispute that they may be treated as property in a generic sense, I need not consider the question further. 
[81] In the proceeding, as I have mentioned, B2C2 had alleged that Quoine’s reversal of the disputed trades was in breach of contract and breach of trust. On the trust point there were no express words in Quoine’s terms and conditions indicating an intention to create a trust. However, B2C2 argued Quoine had shown an intention to create a trust by holding traders’ cryptocurrency in separate digital wallets from Quoine’s own assets. Against that, Quoine submitted that a Risk Disclosure Statement it had provided notified customers that assets were not deposited in a trust account so customers may lose their assets in the case that Quoine was to be bankrupted or go into liquidation. 
[82] The High Court, in considering the first instance claims brought by B2C2, allowed them both on the basis of breach of contract and breach of trust. In finding there was a trust, the Court there held that the “decisive factor” was that the assets were held separately as members’ assets rather than as part of Quoine’s trading assets. The decision was appealed to the Court of Appeal as I have noted. On appeal the majority upheld the High Court’s decision on the breach of contract aspect but overturned the decision on the breach of trust cause of action. On that breach of trust claim, a majority of the Court of Appeal rejected the International Judge’s view that it was a “decisive factor” that the assets were held separately rather than as part of Quoine’s trading assets. The Court of Appeal found the mere fact Quoine’s assets were segregated from its customers cannot in and of itself lead to the conclusion that there was a trust. Further discussion of this trust aspect will follow later in my judgment. 
[83] On the “property” question, in its decision, the Court of Appeal also declined to decide whether Bitcoin as the cryptocurrency in question was “property” capable of forming the subject matter of a trust. In their decision the Court of Appeal in the majority judgment, delivered by Menon CJ, commented: There may be much to commend the view that cryptocurrencies should be capable of assimilation into the general concepts of property. There are, however, different questions as to the type of property that is involved. It is not necessary for us to come to a final position on this question in the present case. 
[84] This comment from the Court of Appeal, although not definitive, along with similar suggestions from other authorities, in my view, are of some help when considering this question as to whether the digital assets here could be regarded as “property”. 
Other authorities 
[85] A second case perhaps supporting this interpretation is a 2018 decision in Vorotyntseva v Money-4 Ltd.  There, Birss J sitting in the Chancery Division of the English High Court granted ex parte a proprietary freezing order over some bitcoin and ethereum currency, stating that the defendant in that case had not suggested that “cryptocurrency cannot be a form of ‘property’”.  No further discussion took place on the point. 
[86] In a not dissimilar Canadian decision, Shair.Com Global Digital Services Ltd v Arnold, the Supreme Court of British Colombia granted an ex parte preservation order to the plaintiff company against its former chief operating officer with respect to digital currencies that might still be in the defendant’s possession.  Without providing any reasoning the Court accepted that cryptocurrencies could be property within the rules for preservation orders, noting that in the correspondence between the parties that had been filed for the proceeding the defendant had not denied that the plaintiff had an interest to pursue. 
[87] Recently, a decision of the English High Court in AA v Persons Unknown also held that cryptocurrencies are “property”.   There, Bryan J granted an interim proprietary injunction against a cryptocurrency exchange over bitcoin which represented proceeds of ransom monies paid out to a hacker by the applicant insurance company. The hackers had installed malware into the insurance company’s computer system, and demanded the company pay a ransom in bitcoin, to regain access to its system. The ransom was paid in bitcoin and transferred into the exchange. The insurance company applied to the Court for an interim proprietary injunction against the exchange over the bitcoin, amongst other things. 
[88] Only counsel for the applicant insurance company appeared at the hearing in that case and filed submissions. And, it seems the High Court there primarily relied on the Legal Statement on Cryptoassets and Smart Contracts, and that no other argument was addressed to the Court on the issue. 
[89] It is also useful, as I see it, to turn to consider a diverse range of types of assets that have already been recognised elsewhere as “property” at equity. These examples of “property” also illustrate that they are capable of being the subject of a trust. They include:
(a) Any simple chose in action– even an oral contract can be the subject of an orally created trust with the result that a liquidator of a corporate trustee could not pursue the chose in order to obtain a money judgment for the benefit of unsecured creditors. 
(b) Non-enforceable debt claims – for example a barrister’s claim that fees be paid by the relevant instructing solicitor was recently held in Gwinnutt v George to be part of the property belonging to a bankrupt barrister, even though the barrister had no legally enforceable right to the fees. In the circumstances of that case, in fact there was also no contract all between the barrister and the solicitors. 
(c) Payments through the banking system – money transactions have recently ceased to involve tangible coins or banknotes and usually take the form of electronic bank payments. Equity will apply its proprietary tracing rules to payments effected by these means, even though on transfer of money from one bank account to another this does not involve the transfer of anything in the literal sense from the payer to the payee and the recipient does not hold the same asset.     
(d) Copyright – although copyright has statutory recognition, it nevertheless provides a strong example of intangible property. The subject matter of copyright turns merely on combinations of sounds or shapes in two or three dimensions (including words or drawings) that are sufficiently distinctive to justify the law preventing others from reproducing them.  These sounds and shapes can exist in digital form. Although the resulting intellectual property needs to be identifiable, in many cases whether there has been a copyright infringement will involve an element of judgment in the tribunal called upon to adjudicate on the associated legal rights. These rights can be made the subject matter of a trust. 
(e) Shares – shares in a company are another type of intangible property which typically has a more complicated existence than merely conferring a right to sue. Voting rights in relation to the appointment and removal of directors and in relation to other important company matters can be exercised. Shares are properly regarded as an item of property in equity even where they are non-transferrable or transferrable only to particular persons.   
(f) Licences/exemptions/quotas – modern statutory regulation frequently operated on the basis of blanket prohibitions coupled with defined exemptions granted to individuals that allow each individual then to trade. Such exemptions function and are recognised as intangible items as property. Their value is not derived from a right to sue but rather the opposite, namely an immunity from prosecution.  Examples of this include export quotas, milk supply quotas, fishing quotas, petroleum exploration licences, waste disposal licences, and carbon credits. These tradeable rights can form the subject matter of a trust and where that happens the asset falls outside the estate of an insolvent trustee. There is a large body of case law that confirms such rights are a type of property and subject to normal property protections.   
(g) A trustee’s rights of indemnity – a trustee’s rights to be indemnified in respect of trust expenses has been held to confer a proprietary interest in the trust assets even though these assets are realised by self-help remedies rather than recourse to the courts: Carter Holt Harvey Woodproducts Australia Pty Ltd v Commonwealth.  
Although these rights are not choses in action, they are a species of intangible property. The breadth of the sort of interests that may be the subject of a trust is also confirmed here at [84] as follows: 
 To describe [the right of indemnity] as constituting a beneficial interest in the trust assets, and so as property, thus acknowledges the characteristic blending of personal rights and obligations with proprietary interests which is the “genius” of the trust institution. Such a beneficial interest falls naturally and ordinarily within the definition of “property” in s 9 of the Corporations Act. 
Although a number of the examples outlined above do involve statutory licences and quotas and are within broad statutory definitions of the word “property” in the respective jurisdictions, the types of interest capable of forming the subject matter of a trust at equity, as I see it, are no less broad. A similar point was made by Mr Stephen Morris QC sitting as a deputy High Court judge in Armstrong DLW GmbH v Winnington Networks Ltd:  Whilst the cited case law concerned the meaning of “property” as specifically defined in various statutes, in my judgment, the reasoning of Morritt LJ (in Celtic Extraction) applies equally to the characteristics of property at common law. Indeed, Morritt LJ himself relied upon National Provincial Bank v Ainsworth.  Moreover the terms used in statutory definitions are themselves derived from common law concepts. 
[90] At this point it is useful also to interpolate three recent New Zealand cases which might be seen to be at the boundaries of the legal concept of “property”. The first is Dixon v R.  In this case, which adopted a broad approach to the concept, the Supreme Court held that a digital copy of CCTV footage was “property” within the broad definition found in s 2 of the Crimes Act 1961. The defendant had downloaded a copy of certain footage without the consent of the owner of the computer on which the footage had been recorded. The Court held that computer data can be “property” and that making a copy of it involves a taking, even when the data is not protected by a password. The Supreme Court appeared to endorse the view that computer data would meet general definitions of property including that within s 4 of the Property Law Act 2007. Arnold J, writing the judgment of the Court, stated :  
We consider that interpreting the word “property” as we have is not only required by the statutory purpose and context but is also consistent with the common conception of “property”. 
[91] The second case, a decision of Thomas J in the High Court is Henderson v Walker.  In that case Thomas J was prepared to apply the principles of Dixon in a private law setting and to extend the tort of conversion to purely personal digital information, including the content of private emails. However, her Honour also concluded merely making a copy of emails and other personal data would not amount to conversion. Refusing access to them or destroying them would be, nevertheless. 
[92] In that case, Mr Walker in his capacity as liquidator of subsidiary companies of Property Ventures Limited (PVL) came into possession of a laptop belonging to PVL and of a tape drive that was a backup of PVL’s server. There were a lot of personal, non-company emails sent by and to Mr Henderson, a principal of PVL, and some personal photographs on those devices. Mr Walker distributed at least some of these or allowed them to be distributed to third parties who should not have received this material. Mr Henderson sued Mr Walker, pleading some seven causes of action including breach of confidence, invasion of privacy and conversion. Thomas J held that in principle the common law action in conversion was available with respect to some of the actions which had occurred involving the computer data. 
[93] In my view, it is reasonable to conclude that the reasoning of Thomas J that this data was effectively “property” capable of being converted, could be properly extended to wrongful interferences with cryptocurrency or digital assets. Any person who gained unauthorised access to the private key attached to cryptocoins and used it would permanently deprive the proper possessor of the cryptocoins of that property and its value. 
[94] Another recent High Court decision in New Zealand, Commissioner of Police v Rowland, is also usefully noted here.  In that case this Court approved a settlement under the Criminal Proceeds (Recovery) Act 2009 that included quantities of two cryptocurrencies – bitcoin and ethereum. The question whether the cryptocurrencies were “property” that was amenable to forfeiture under that legislation, however, was not raised in the proceeding. An assumption was made that they did fall within the definition in terms of that legislation. The definition of “property” in the Criminal Proceeds (Recovery) Act at s 5 provides: property— (a) means real or personal property of any kind— (i) whether situated in New Zealand or a foreign country; and (ii) whether tangible or intangible; and (iii) whether movable or immovable; and (b) includes an interest in real or personal property 
[95] Turning back to the decisions noted above in Dixon and Henderson, in those cases the New Zealand courts involved have accepted that the orthodox position that information is not “property” does not attach to cases involving digital assets. There, digital files were seen as “property” by distinguishing them from “pure information”. 
[96] So far as the Supreme Court was concerned in Dixon v R, in the context of the Crimes Act 1961, this was because the files (the digital footage) there: (a) could be identified; (b) had a value; (c) were capable of being transferred; and (d) had a physical presence, albeit one that could not be detected by means of unaided sensors. 
[97] In Thomas J’s decision in this Court in Henderson, in the context of the tort of conversion, this was because it was possible to control and therefore possess the digital files (a large number of documents, emails and images). Possession required cognitive control and manual control. While traditionally the tort of conversion requires physical control and therefore tangibility, physical control is only one example of manual control. The two fundamental elements of manual control are excludability and exhaustibility – whether others can be excluded from the thing’s control and when the thing’s value can be deprived from others. In her decision Thomas J considered both were satisfied on the facts because: 
(a) As to excludability: digital files have a material presence. They physically alter the medium on which they are held. The physical presence allows others to be excluded from the digital asset, either by physical control of the medium or by password protection. 
(b) As to exhaustibility: digital files can be deleted or modified so as to render them useless or inaccessible. 
[98] These principles, in my view, apply equally in the present case to the cryptocurrencies at issue. 
[99] I turn now to the Companies Act. In that Act reference is made to both “property” and “assets”. Assets are not defined in the Act other than the section specific definition at s 129 which applies to “major transactions”. Section 129(2) provides: ...assets includes property of any kind, whether tangible or intangible That definition is expressly limited to s 129 and the use of inclusive language supports the finding that the term “asset” might possibly be seen as wider in scope than “property”. 
[100] The powers of liquidators in the Act are generally expressed to be over a company’s “assets”:
(a) Section 248(1)(a) provides that: The liquidator has custody and control over the company’s assets. 
(b) Section 253 characterises the principal duty of a liquidator as: (a) to take possession of, protect, realise and distribute the assets, or the proceeds of the realisation of the assets, of the company to its creditors in accordance with the act; and (b) if there are surplus assets remaining, to distribute them, or the proceeds of the realisation of the surplus assets, in accordance with s 313(4) in a reasonable and efficient manner. 
[101] The term “asset” is used elsewhere in the Companies Act: 
(a) The solvency test: the relevant limb of the test here is that: “the value of the company’s assets is greater than the value of its liabilities...”. 
(b) Section 237 provides that the Court may make additional orders relating to (among other things): ... (a) the transfer or vesting of real or personal property, assets, rights, powers, interests, liabilities, contracts, and engagements: ... 
(c) Clause 1(1) of sch 7 requires the liquidator to pay: ... 
(e) to any creditor who protects, preserves the value of, or recovers assets of the company for the benefit of the company’s creditors by the payment of money or the giving of an indemnity,— (i) the amount received by the liquidator by the realisation of those assets, up to the value of that creditor’s unsecured debt; and (ii) the amount of the costs incurred by that creditor in protecting, preserving the value of, or recovering those assets. 
The four requirements for a “property” interest 
[102] I return now to the classic statement of the characteristics of “property” outlined by Lord Wilberforce in Ainsworth essentially to recognise what constitutes a “property” interest, and then to apply this to each cryptocurrency at issue here.  In doing so, I need to say at the outset that I am satisfied the criteria for Lord Wilberforce’s definition of “property” are clearly met in this case. I say this bearing in mind the indications I have outlined from the range of authorities noted above that support this conclusion. This is also in line with the approach adopted in the Legal Statement on Cryptoassets and Smart Contracts noted above. 
[103] Lord Wilberforce’s long-applied statement is outlined at [75] above. It outlines four requirements that I now address in turn. 
(a) Identifiable subject matter 
[104] The first requirement is that the asset in question needs to be definable. It needs to be capable of being isolated from other assets whether of the same type or of other types and thereby identified. It is possible, however, for there to be co-ownership (either at law or in equity) of a definable share of an identified bulk of like assets. The present situation, as I see it, is one of this sort. 
[105] Computer-readable strings of characters recorded on networks of computers established for the purpose of recording those strings, as I see it, are sufficiently distinct to be capable of then being allocated uniquely to an accountholder on that particular network. For the cryptocurrencies involved here, the allocation is made by what is called a public key – the data allocated to one public key will not be confused with another. This is the case even though the identical data is held on every computer attached to the network. Indeed, the working of the system is such that the distribution of the data across a large network of computers, when combined with cryptography that prevents individual networks from altering historic data over the network, assists in giving that data stability. It is these features that provide the basic underpinning for the existing cryptocurrencies. 
[106] This is in large measure similar to what occurs in the banking system where large and trusted international banks record balances in various numbered bank accounts held with them. The identifiability provided by cryptocurrency data recorded in the network of computers (called the “distributed ledger”) is no less than the identifiability which results from the bank’s inclusion of balances in their customers’ numbered bank accounts. Equity regards such recorded bank balances as a type of property owned by the party in whose favour the balance is recorded. 
[107] The developer of the most widely known cryptocurrency (Bitcoin), Satoshi Nakamoto, who I have referred to above, argued in 2008:
...an electronic payment system based on cryptographic proof instead of trust, allowing any two willing parties to transact directly with each other without the need for a trusted third party... will provide superior stability and reliability compared to the traditional banking system. 
[108] It is also the case, as I see it, that the public key so allocated to a cryptocurrency account might also be argued to be more readily identifiable than some asserted rights for example to copyright (which is acknowledged as “property”) where issues of originality may be at play. 
(b) Identifiable by third parties 
[109] The second component of property outlined by Lord Wilberforce is that the thing needs to be identifiable by third parties. This element alludes to the thing identified having to have an owner capable of being recognised as such by third parties. The degree of control over the type of asset that a person has to have before the law recognises it as capable of being owned must involve an element of judgement but again I am satisfied here that cryptoassets clearly meet this criterion. 
[110] On this aspect, it has long been recognised by property lawyers that the power of an owner to exclude others from an asset provides a more important indicator of ownership than the power actively to use or benefit from that asset. 
[111] The unique strings of data recording the creation of and dealings with cryptocurrency are always allocated via the public key to a particular accountholder connected to the system. But that allocation by itself is unlikely to be recognised as creating an item of property if there is no element of excludability. So, if that accountholder’s personal connection to the data via the public key could be lost through any person connected with the network being able to reallocate the cryptocurrency to any other colleague on the network without the consent of the accountholder, there might be some doubt whether the law would conclude that the accountholder owned the key. 
[112] The degree of control necessary for ownership (namely the power to exclude others) is achieved for cryptocurrencies by the computer software allocating to each public key a second set of data made available only to the holder of the account (the private key), and requiring the combination of the two sets of data in order to record a transfer of the cryptocurrency attached to the public key from one account to another. A varied public key and a new private key for the cryptocurrency are generated after each transfer of cryptocurrency. The private key, in effect, is like a PIN. Anyone who learns of the private key attached to a public key can transfer the public key but the private key, having been used once in respect of the public key, cannot be used again. 
[113] These features of cryptocurrencies inhibit two potential practices. First, the existence of the private key inhibits the possibility of involuntary transfers – it gives the power to exclude third parties from access. And secondly, the creation of a new private key after each transfer or disposition inhibits a holder from purporting to transfer the cryptocurrency data twice. 
(c) Capable of assumption by third parties 
[114] The third of Lord Wilberforce’s criteria, namely that the right or interest in question must be capable of assumption by third parties, generally involves two aspects: 
(a) Third parties must respect the rights of the owner in that property and will be subject to actions expressly devised by the law to give effect to proprietary rights if they assert their own claim to ownership without justification. Property has been said by its nature to be concerned with legal rights that affect strangers to bilateral transactions.  These third parties will also include insolvency officials of an insolvent trustee; and 
(b) Normally, but not always, an asset recognised by the law as an item of property will be something which is potentially desirable to third parties such that they would want themselves to obtain ownership of it. It might not matter that an asset has no current market value if there has been a market for the asset in the past. For example, in the case where polluted land has excessive clean-up costs, it may be worthless, but it will still be regarded as property. 
[115] Both aspects of this component of Lord Wilberforce’s test are reflected in comments by Lord Bridge for the Privy Council in Attorney-General of Hong Kong v Nai-Keung, a case concerned with a charge of theft of an export quota brought under the theft ordinance of Hong Kong:  It would be strange indeed if something which is freely brought and sold and which may clearly be the subject of dishonest dealing which deprives the owner of the benefit it confers were not capable of being stolen. Their Lordships have no hesitation in concluding that export quotas in Hong Kong, although not “things in action” are a form of “other intangible property. 
[116] I am satisfied here that cryptocurrencies meet both aspects of the assumption by third parties criterion outlined by Lord Wilberforce. There can be no doubt that cryptocurrencies can be, and many are, the subject of active trading markets. 
(d) Some degree of permanence or stability 
[117] The last of Lord Wilberforce’s criteria for determining whether something is capable of attracting proprietary status, in my view is also met here. This criterion requires that the thing needs to have some degree of permanence or stability, but as I see the position, it does not add much to the other three criteria noted above. It is true too that some assets will have little permanence yet undoubtedly be property, such as the example of the ticket to a football match which can have a very short life yet unquestionably it is regarded as property. Also unproblematic, as I see it, will be situations where the short life of an asset is the result of the deliberate process of transferring the value inherent in the asset so that one asset becomes replaced by another. As I have noted above, cryptocurrencies work in this manner but it is also true that bank payments use a similar process which is simply native to the type of property in question. This is not inimical to the asset’s status as property. 
[118] The blockchain methodology which cryptocurrency systems deploy also greatly assist in giving stability to cryptocoins. The entire life history of a cryptocoin is available in the public recordkeeping of the blockchain. A particular cryptocoin stays fully recognised, in existence and stable unless and until it is “spent” through the use of the private key, which may never happen. Standard cryptocurrency systems do not provide for the arbitrary cancellation of coins. 
[119] While it is possible for cryptocurrencies to be wrongfully interfered with, by someone gaining unauthorised access to the private key or by hacking the address to which an owner intends to send a coin, these risks are not markedly greater than those borne by an owner of tangible property or a person relying on the integrity of a bank account record with or without the use of a PIN. 
Conclusion on the four criteria 
[120] I am satisfied that cryptocurrencies meet the standard criteria outlined by Lord Wilberforce to be considered a species of “property”. They are a type of intangible property as a result of the combination of three interdependent features. They obtain their definition as a result of the public key recording the unit of currency. The control and stability necessary to ownership and for creating a market in the coins are provided by the other two features – the private key attached to the corresponding public key and the generation of a fresh private key upon a transfer of the relevant coin. 
[121] This identical point is made in the Legal Statement on Cryptoassets and Smart Contracts which says that a cryptoasset is “a conglomeration of public data, private key and system rules.” 
Possible arguments against cryptocurrency being property 
[122] Two arguments that are most commonly raised to suggest that cryptocurrencies do not have the status of “property” are: 
(a) The common law recognises only two classes of personal property: tangibles and choses in action. Cryptocurrencies are said to be neither. 
(b) Information is not generally recognised as a form of “property” and cryptocurrencies might be said to be a form of information. 
[123] Although before me counsel for the creditors did not rely particularly on the first objection noted above, nevertheless I address it briefly. On this, I am satisfied the argument here is in fact a red-herring. This is because cases which might be perceived to be problematic in this area are not about the limits of what can be recognised as “property” but simply about the number of categories of “property” one needs. This accords with the well-known dictum of Fry LJ sitting in the English Court of Appeal in Colonial Bank v Whinney that all personal property must either be a chose in possession or a chose in action. The argument follows that cryptocurrencies are neither a chose in possession nor a chose in action. 
[124] Essentially here, Fry LJ in his judgment did not seem to be taking a narrow view of what can be classified as property, but rather he was simply wanting to push all examples of property into one of two categories. There is nothing, as I see it, in Fry LJ’s dictum that would lead a court to conclude that cryptocurrencies are not property. The most that could be said is that cryptocoins might have to be classified as choses in action. Indeed, it would be ironic that something that might be said to have more proprietary features than a simple debt is deemed not to be property at all when a simple debt qualifies. 
[125] For these reasons, this first argument advanced by some to support the claim that cryptocurrency is not property in my view is readily dismissed. 
[126] I turn now to the second argument suggesting that cryptocurrency does not have the status of property as noted at [122](b) above. This is to the effect that cryptocoins are just a type of information and that information is not property. The argument is based on the view that neither the common law nor equity recognises property in “information” and cryptocurrencies are said to be merely digitally recorded information. This argument, it is said, is supported by the 2014 decision of the English Court of Appeal in Your Response Ltd v Datateam Business Media Ltd. In Your Response, the Court held that there could be no property in a database in the situation prevailing there, which involved a party contracted by a client to maintain and update a database of the client’s customers. It was held that this party had no common law lien over the database for the fees owed to it. As I see it, however, the decision in Your Response does not go much further than to make a determination upon the particular facts of that case. I am satisfied it is an inconclusive precedent in a case such as the present. 
[127] And, in my view, it is wrong in any event to regard cryptocurrencies as mere information because: 
(a) The whole purpose behind cryptocurrencies is to create an item of tradeable value not simply to record or to impart in confidence knowledge or information. Although cryptocoins are not backed by the promise of a bank, the combination of data that records their existence and affords them exclusivity is otherwise comparable to the electronic records of a bank. The use of the private key also provides a method of transferring that value. This might be seen as similar in operation to, for example, a PIN on an electronic bank account. 
(b) And, generally, as I see it, cryptocoins are no more mere information than the words of a contract are. What allows a contract to be capable of being an item of property is not the words nor even the binding promise which is only a personal obligation, but the fact that equity recognises there is a unique relationship between the parties created by the words and then supplies a system for transferring the contractual rights. Similarly, a unique relationship and system of transfer exists with respect to the relevant data on the blockchain that makes up a cryptocoin. 
(c) In Boardman v Phipps Lord Upjohn stated: "In general, information is not property at all. It is normally open to all who have eyes to read and ears to hear.” This statement appears to confirm as a principle for not regarding information as property the fact that it can be infinitely duplicated. Again, this is not true of cryptocoins where every public key recording the data constituting the coin is unique on the system where it is recorded. It is also protected by the associated private key from being transferred without consent. 
(d) Cryptocurrency systems provide a more secure method of transfer than a mere assignment of a chose in action. It is possible in equity for the holder of a chose in action to assign it multiple times. Only one assignment will be effective to bind the debtor but the winner may not be the first assignee in time but rather the first assignee to notify the debtor. By way of contrast, a cryptocoin can not only be assigned in that way but it can also be sold only once. 
[128] I am satisfied that cryptocurrencies are far more than merely digitally recorded information. The argument that cryptocurrency is mere information and therefore it is not property is a simplistic one and, in my view, it is wrong in the present context. I dismiss it. 
Public policy arguments 
[129] Lastly, I turn to certain public policy arguments here. It is widely known that at least some types of cryptocurrency are used by criminals for the transmission of funds across borders in order to pursue criminal activity and as a means of laundering the proceeds of past criminal activity. This is not exclusive, however. Cryptocurrencies have also become popular with honest people as a method of effecting payments and of investing. The traditional banking sector is itself widely reported to be already using block chain technology and to be planning to create trading platforms for cryptocurrencies.  Any failure by the general law to recognise cryptocurrencies as property, as I see it, would have little effect in reducing potential criminal activity. The banking system is subject to exploitation by the criminal fraternity just as other traditional assets are. 
[130] In my view, honest commercial developments may very well be hindered by a failure of the general law to recognise cryptoassets as property. This is notwithstanding any possible need for more formal regulation of cryptocurrencies. 
[131] The Legal Statement on Cryptoassets and Smart Contracts has also advocated dealing with the status of cryptocurrencies unencumbered by other legal issues including the need for regulation.  Similarly, in those cases where the status of cryptocurrencies as property has been assumed or conceded, including those I have noted above, no court has felt obliged to take a public policy objection. Further, before me Ms Cooper for the creditors raised no particular public policy arguments. 
[132] Overall, I am of the view that public policy questions here do nothing to harm the accountholders’ contention that cryptocurrencies do have the status of property. 
Conclusion 
[133] The answer to the question posed at [46](a) above is yes. I find that, for the reasons outlined above, all of the various cryptocurrencies are “property” within the definition outlined in s 2 of the Companies Act and also probably more generally. In addition, these digital assets, I find, being property, are capable of forming the subject matter of a trust.
'Cryptocurrencies as Property' by Paul T Babie, David Brown, Ryan Catterwell and Mark Giancaspro comments
The case provides significant guidance for any jurisdiction, common or civil, faced with determining whether cyrptocurrencies are property. This note outlines the approach taken to ‘the property question’ by Gendall J, in four parts. Part I introduces the property question. Part II provides a brief overview of blockchain and the nature of cryptocurrencies. Part III briefly recounts Gendall J’s reasons for the judgment
In the United Kingdom  AA v Persons Unknown and Others, Re Bitcoin  [2019] EWHC 3556 (Comm) the English Commercial Court has concluded that Bitcoin is property.

The judgment states, at [55] onwards -
Turning then to the relevant principles in relation to the granting of a proprietary injunction, the first and perhaps fundamental question that arises in relation to this claim for a proprietary injunction is whether or not in fact the Bitcoins, which are being held in this account of the second defendant with the third or fourth defendants are property at all. Prima facie there is a difficulty in treating Bitcoins and other crypto currencies as a form of property: they are neither chose in possession nor are they chose in action. They are not choses in possession because they are virtual, they are not tangible, they cannot be possessed. They are not choses in action because they do not embody any right capable of being enforced by action. That produces a difficulty because English law traditionally views property as being of only two kinds, choses in possession and choses in action. In Colonial Bank v Whinney [1885] 30 Ch.D 261 Fry LJ said: "All personal things are either in possession or action. The law knows no tertium quid between the two." 
On that analysis Bitcoins and other crypto currencies could not be classified as a form of property, which would prevent them being the subject of a proprietary injunction or a freezing injunction. This exact issue has recently in November 2019 been the subject of detailed consideration by the UK Jurisdictional Task Force ("UKJT") which has published a legal statement on Crypto assets and Smart contracts, ("the Legal Statement"). The UKJT is chaired by Sir Geoffrey Vos, and Sir Antony Zacaroli is also a member. However, neither in their judicial capacity was responsible for the drafting of the legal statement, nor have either in their judicial capacities endorsed that legal statement. Indeed Sir Geoffrey Voss explained in the foreword to the Legal Statement: "It is not my role as a judge nor that of the UKJT or its parent, the UK Lawtech Delivery Panel, to endorse the contents of the Legal Statement". Those responsible for drafting the Legal Statement were Laurence Akka QC, David Quest QC, Matthew Lavy and Sam Goodman. 
It follows that the legal statement is not in fact a statement of the law. Nevertheless, in my judgment, it is relevant to consider the analysis in that Legal Statement as to the proprietary status of crypto currencies because it is a detailed and careful consideration and, as I shall come on to, I consider that that analysis as to the proprietary status of crypto currencies is compelling and for the reasons identified therein should be adopted by this court. 
The difficulty identified in treating crypto currencies in property, as I say, starts from the premise that the English law of property recognises no forms of property other than choses in possession and choses in action. As I have already identified, crypto currencies do not sit neatly within either category. However, on a more detailed analysis I consider that it is fallacious to proceed on the basis that the English law of property recognises no forms of property other than choses in possession and choses in action. The reasons for this are set out between paragraphs 71 to 84 in the Legal Statement.
"71. The Colonial Bank case concerned a dispute about shares deposited as security for a loan. The borrower was declared bankrupt and there was a contest for the shares between the plaintiff bank and the trustee in bankruptcy. The case was not about the scope of property generally: there was no dispute that the shares were property. The relevant question was rather whether they were things in action within the meaning of the Bankruptcy Act 1883, an issue of statutory interpretation. If so, then they were excluded from the bankrupt estate by section 44 of that Act. 
72. Lindley LJ and Cotton LJ held that the shares were not things in action. They relied principally on previous case law where the court had come to a similar conclusion in relation to the predecessor statute, the Bankruptcy Act 1869. They also drew some support from sections 50(3) and 50(5) of the 1883 Act, which appeared to make a distinction between shares and things in action. 
73. Fry LJ reached the opposite conclusion, reasoning principally from what he considered to be the essential nature of a share. A share constituted "the right to receive certain benefits from a corporation, and to do certain acts as a member of that corporation" and was therefore, in his view, closely akin to a debt. He supported his conclusion by a comparison of shares to other, established, things in action, such as partnership interests and interests in funds. 
74. Fry LJ's statement that "personal things" are either in possession or in action, and that there is no third category, may carry the logical implication that an intangible thing is not property if it is not a thing in action. It is not clear, however, whether Fry LJ intended that corollary and it should not in any case be regarded as part of the reasoning leading to his decision (and so binding in other cases). The question before him was whether the shares were things in action for the purpose of the Bankruptcy Act, not whether they were property, still less the scope of property generally. 
75. Moreover, in making the statement Fry LJ attributed a very broad meaning to things in action. He approved a passage from Personal Property by Joshua Williams, which described things in action as a kind of residual category of property: "In modern times [sc. by the 19th century] … several species of property have sprung up which were unknown to the common law … For want of a better classification, these subjects of personal property are now usually spoken of as ... [things] in action. They are, in fact, personal property of an incorporeal nature…". 
76. On appeal, the House of Lords also framed the question as one about statutory interpretation. They reversed the Court of Appeal's decision, approving the judgment and reasoning of Fry LJ. They did not explicitly address the issue of exhaustive classification between things in action and things in possession and said nothing about the definition of property. Lord Blackburn did say, however, that "in modern times lawyers have accurately or inaccurately used the phrase '[things] in action' as including all personal chattels that are not in possession". Thus, to the extent that the House of Lords agreed with Fry LJ on the classification issue, that seems to have been on the basis that the class of things in action could be extended to all intangible property (i.e. it was a residual class of all things not in possession) rather than on the basis that the class of intangible property should be restricted to rights that could be claimed or enforced by action. 
77. Our view is that Colonial Bank is not therefore to be treated as limiting the scope of what kinds of things can be property in law. If anything, it shows the ability of the common law to stretch traditional definitions and concepts to adapt to new business practices (in that case the development of shares in companies). 
78. Colonial Bank was referred to in Allgemeine Versicherungs-Gesellschaft Helvetia v Administrator of German Property by Slesser LJ as showing "how the two conditions of [thing] in action and [thing] in possession are antithetical and how there is nomiddle term". Again, however, the case was not about the scope of property generally but about whether something that was undoubtedly property should be classified as a thing in possession or a thing in action. 
79. Most recently, Colonial Bank was cited in 2014 in Your Response v Datateam. In that case, the claimant sought to assert a lien over a database in digital form but faced the obstacle of the previous decision of the House of Lords in OBG Ltd v Allan that there could be no claim in conversion for wrongful interference with a thing in action because it could not be possessed. In an attempt to distinguish the case from OBG, the claimant argued that, even if the database could not be regarded as a physical object, it was a form of intangible property different from a thing in action and so was capable of being possessed. 
80. The Court of Appeal rejected the argument. Moore-Bick LJ said that Colonial Bank made it "very difficult to accept that the common law recognises the existence of intangible property other than [things] in action (apart from patents, which are subject to statutory classification), but even if it does, the decision in OBG Ltd v Allan [2008] AC 1 prevents us from holding that property of that kind is susceptible of possession so that wrongful interference can constitute the tort of conversion." He said that there was "a powerful case for reconsidering the dichotomy between [things] in possession and [things] in action and recognising a third category of intangible property, which may also be susceptible of possession and therefore amenable to the tort of conversion" but the Court of Appeal could not do that because it was bound to follow the decision in OBG. The other members of the court agreed. 
81. The Court of Appeal did not, and did not need to, go so far as to hold that intangible things other than things in action could never be property at all, only that they could not be the subject of certain remedies. The intangible thing with which they were concerned was a database, which (as Floyd LJ said) would not be regarded as property anyway because it was pure information. They did not have to consider intangible assets with the special characteristics possessed by cryptoassets. 
82. In other cases, the courts have found no difficulty in treating novel kinds of intangible assets as property. Although some of those cases are concerned with the meaning of property in particular statutory contexts, there are at least two concerning property in general. In Dairy Swift v Dairywise Farms Ltd, the court held that a milk quota could be the subject of a trust; and in Armstrong v Winnington, the court held that an EU carbon emissions allowance could be the subject of a tracing claim as a form of "other intangible property", even though it was neither a thing in possession nor a thing in action. 
83. A number of important 20th century statutes define property in terms that assume that intangible property is not limited to things in action. The Theft Act 1968, the Proceeds of Crime Act 2002, and the Fraud Act 2006 all define property as including things in action "and other intangible property". It might be said that those statutes are extending the definition of property for their own, special purposes, but they at least demonstrate that there is no conceptual difficulty in treating intangible things as property even if they may not be things in action. Moreover, the Patents Act 1977 goes further in providing, at s30, that a patent or application for a patent "is personal property (without being a thing in action)". That necessarily recognises that personal property can include things other than things in possession (which a patent clearly is not) and things in action. 
84. We conclude that the fact that a cryptoasset might not be a thing in action on the narrower definition of that term does not in itself mean that it cannot be treated as property." 
The conclusion that was expressed was that a crypto asset might not be a thing in action on a narrow definition of that term, but that does not mean that it cannot be treated as property. Essentially, and for the reasons identified in that legal statement, I consider that a crypto asset such as Bitcoin are property. They meet the four criteria set out in Lord Wilberforce's classic definition of property in National Provincial Bank v Ainsworth [1965] 1 AC 1175 as being definable, identifiable by third parties, capable in their nature of assumption by third parties, and having some degree of permanence. That too, was the conclusion of the Singapore International Commercial Court in B2C2 Limited v Quoine PTC Limited [2019] SGHC (I) 03 [142]. 
There are also two English authorities to which my attention has been drawn where crypto currencies have been treated as property, albeit that those authorities do not consider the issue in depth. They are, and I have already mentioned them, in Vorotyntseva v Money -4 Limited t/a as Nebeus .com, the decision of Birss J, where he granted a worldwide freezing order in respect of a substantial quantity of Bitcoin and Ethereum, another virtual currency, and the case of Liam David Robertson, where Moulder J granted an asset preservation order over crypto currencies in that case. 
In those circumstances and for the reasons I have given, as elaborated upon in the Legal Statement which I gratefully as what I consider to be an accurate statement as to the position under English law, I am satisfied for the purpose of granting an interim injunction in the form of an interim proprietary injunction that crypto currencies are a form of property capable of being the subject of a proprietary injunction.

17 April 2020

Haptics and consent

'Teledildonics and rape by deception' by Robert Sparrow and Lauren Karas in (2020) Law, Innovation and Technology 1-20 comments
It is now possible to buy sex toys that connect to the user’s phone or computer via Bluetooth and can be controlled remotely. The use of such Internet-enabled haptic sex toys involves an ineliminable risk of being deceived about particular features of one’s sexual partner and/or about which person one was having ‘sex’ with. Where this occurs, it is possible that the user would become the victim of rape-by-deception. We argue that determining whether a person using an Internet-enabled haptic sex toy has been raped or not when they are involved in a sexual encounter with someone – or something – other than that they intended requires us to confront difficult questions about the definition and significance of sexual intercourse and about the nature and harm of rape. Our discussion of these topics suggests that the use of such devices is more ethically fraught than has been appreciated to date. 
 The authors argue
The search for new or improved sexual pleasures plays a significant – if often under acknowledged – role in driving the development of new technologies. It was perhaps inevitable, then, that progress in ‘haptics’ – the science and technology of the transmission of touch – would spark interest in the development of haptic sex toys.  Consequently, it is now possible to purchase a number of sex toys that transmit touch and physical sensation via the Internet.   
In this paper, we want to reflect on these technologies and some of the ethical and philosophical questions they raise for two reasons. First, given the popularity of ordinary sex toys, and the sexual opportunities and communities made possible by the Internet, it is reasonable to assume that large numbers of people will experiment with these new remote-controlled and interactive sex toys and that a significant number will use them regularly. Any ethical and/or philosophical issues they raise are thus of interest simply by  virtue of the number of people they might affect. Second, the use of such devices seems to involve a not-insignificant risk of users being deceived about the identity of the person with whom they are having ‘sex’. As we shall see below, it is possible that in such cases the user would become the victim of ‘rape by deception’. Until this issue can be resolved, the design and manufacture of teleoperated and remote-controlled sex toys involves profound moral hazards. 
While existing products fall short of allowing fully immersive ‘cybersex’, it seems likely that in the not-too-distant future devices that transmit a larger range of genital sensations, which we shall call Internet-enabled haptic sex  toys (henceforth IEHSTs), will be developed. In order to bring the philosophical questions that interest us into stark relief and to reduce the risk of our discussion being rendered obsolete by technological progress, we shall for the most part discuss the issues raised by the use of such IEHSTs. We suggest that determining whether a person using an IEHST has been raped or not, when they are involved in a sexual encounter with someone – or some thing – other than that they intended, requires one to confront difficult questions about the definition and significance of sexual penetration, what counts as consent, and the nature and harm of rape. Our discussion of these topics will draw upon the academic literature on the philosophy of sex, philosophical and feminist discussions of the nature of rape, and – in particular – the literature on rape by deception. We argue that if one allows that IEHSTs enable sexual intercourse via the Internet then they will involve a significant risk of rape by deception. This risk implies that the use of such devices would be – and, we suggest, the use of existing devices is – more ethically fraught than has been appreciated to date. We also hope, throughout our discussion, to show how thinking about IEHSTs offers a valuable opportunity to gain new insights into some old questions in the philosophy of sex. 
An important limitation of our discussion is that we are only concerned with the ethical and philosophical issues that are raised by the risk of rape involved in the use of these devices. As will become abundantly clear in the discussion that follows, it seems likely that even if one concludes that these devices do not involve a risk of rape, they do involve a significant risk of sexual assault, which might itself be enough to raise ethical red flags about their design and use. However, because of the length and philosophical complexity of our investigation of the risk of rape involved, this further set of questions must remain a topic for future investigations. 
Relatedly, we have not tried to settle the question of how IEHSTs should be regulated here for a number of reasons. First, as we hope our discussion demonstrates, the ethical and conceptual questions arising from the possibility of deception involved in cybersex are complex and profound. It is, we shall argue, plausible to hold that the use of IEHSTs may expose users to a risk of rape and also make it easier for malicious actors to rape people. However, it also seems likely that these devices will be popular with a class of potential users who do not share the philosophical commitments that suggest that deception in the context of the use of IEHSTs can constitute rape. Before we can decide whether – or how – we should regulate IEHSTs, then, it is important to consider the philosophical questions we address below. Moreover, second, even if one wished to regulate to minimise the risks posed by these devices, the social acceptability, and thus the effectiveness, of such regulation is likely to depend in part on whether the regulations track people’s intuitions about the nature and significance of the wrong done by those who misuse IEHSTs in various ways. Again, then, at a bare minimum we need to know if – and when – the wrong might be rape. Third, developing good regulation in this area would require paying attention to various pragmatic and technical matters (How hard would it be to identify those who hacked into such devices? Would it be possible to ensure that minors could not access them? What means might be available to prosecute people misusing these devices across national borders?) that are beyond the bounds of our expertise. For these reasons, we have chosen to leave the question of appropriate regulation to future investigators, but hope that, by clarifying the underlying ethical and conceptual issues, our own work will make their task easier. 
In the first section of the paper we provide a brief account of the existing range of remote-controlled and interactive sex toys, as well as the likely future of this technology, and explain what we shall understand by IEHSTs for the purposes of the current investigation. The second section of the paper outlines the prima facie case for allowing that the use of IEHSTs would constitute sex rather than masturbation. In the third section, we introduce the idea of ‘rape by deception’ and discuss some of the ways it challenges our intuitions regarding rape in other contexts. The fourth section argues that the use of IEHSTs would involve an ineliminable risk of being deceived about particular features of one’s sexual partner and/or about the identity of the person with whom one was having sex. In the fifth section, we consider a number of hypothetical scenarios designed to draw out the implications of such deception for the use of IEHSTs on the assumption that they do enable individuals who are separated by distance to have sex. The sixth section discusses some possible objections to our treatment of the cases in the previous section. In the seventh section, we assume, for the sake of argument, that the use of IEHSTs only involves masturbation and discuss a number of further hypothetical scenarios intended to draw out the implications of deception on this account. The eighth section considers the question of the reasonableness of beliefs about consent in the context of cybersex involving IEHSTs. We conclude by considering the implications of our discussion for the ethics of the use and design of IEHSTs now and in the future.

16 April 2020

Facial Observation and Emotion Scanning

'The Inconsentability of Facial Surveillance' by Evan Selinger and Woodrow Hartzog in (2019) 66 Loyola Law Review 101 comments
 Governments and companies often use consent to justify the use of facial recognition technologies for surveillance. Many proposals for regulating facial recognition technology incorporate consent rules as a way to protect those faces that are being tagged and tracked. But consent is a broken regulatory mechanism for facial surveillance. The individual risks of facial surveillance are impossibly opaque, and our collective autonomy and obscurity interests aren’t captured or served by individual decisions. 
In this article, we argue that facial recognition technologies have a massive and likely fatal consent problem. We reconstruct some of Nancy Kim’s fundamental claims in Consentability: Consent and Its Limits, emphasizing how her consentability framework grants foundational priority to individual and social autonomy, integrates empirical insights into cognitive limitations that significantly impact the quality of human decision-making when granting consent, and identifies social, psychological, and legal impediments that allow the pace and negative consequences of innovation to outstrip the protections of legal regulation. 
We also expand upon Kim’s analysis by arguing that valid consent cannot be given for face surveillance. Even if valid individual consent to face surveillance was possible, permission for such surveillance is in irresolvable conflict with our collective autonomy and obscurity interests. Additionally, there is good reason to be skeptical of consent as the justification for any use of facial recognition technology, including facial characterization, verification, and identification.

'Emotional Expressions Reconsidered: Challenges to Inferring Emotion From Human Facial Movements' by Lisa Feldman Barrett, Ralph Adolphs, Stacy Marsella, Aleix M. Martinez and Seth D Pollak in (2019) 20(1) Psychological Science in the Public Interest comments 

It is commonly assumed that a person’s emotional state can be readily inferred from his or her facial movements, typically called emotional expressions or facial expressions. This assumption influences legal judgments, policy decisions, national security protocols, and educational practices; guides the diagnosis and treatment of psychiatric illness, as well as the development of commercial applications; and pervades everyday social interactions as well as research in other scientific fields such as artificial intelligence, neuroscience, and computer vision. In this article, we survey examples of this widespread assumption, which we refer to as the common view, and we then examine the scientific evidence that tests this view, focusing on the six most popular emotion categories used by consumers of emotion research: anger, disgust, fear, happiness, sadness, and surprise. The available scientific evidence suggests that people do sometimes smile when happy, frown when sad, scowl when angry, and so on, as proposed by the common view, more than what would be expected by chance. Yet how people communicate anger, disgust, fear, happiness, sadness, and surprise varies substantially across cultures, situations, and even across people within a single situation. Furthermore, similar configurations of facial movements variably express instances of more than one emotion category. In fact, a given configuration of facial movements, such as a scowl, often communicates something other than an emotional state. Scientists agree that facial movements convey a range of information and are important for social communication, emotional or otherwise. But our review suggests an urgent need for research that examines how people actually move their faces to express emotions and other social information in the variety of contexts that make up everyday life, as well as careful study of the mechanisms by which people perceive instances of emotion in one another. We make specific research recommendations that will yield a more valid picture of how people move their faces to express emotions and how they infer emotional meaning from facial movements in situations of everyday life. This research is crucial to provide consumers of emotion research with the translational information they require.

The authors argue 

Faces are a ubiquitous part of everyday life for humans. People greet each other with smiles or nods. They have face-to-face conversations on a daily basis, whether in person or via computers. They capture faces with smartphones and tablets, exchanging photos of themselves and of each other on Instagram, Snapchat, and other social-media platforms. The ability to perceive faces is one of the first capacities to emerge after birth: An infant begins to perceive faces within the first few days of life, equipped with a preference for face-like arrangements that allows the brain to wire itself, with experience, to become expert at perceiving faces (Arcaro, Schade, Vincent, Ponce, & Livingstone, 2017; Cassia, Turati, & Simion, 2004; Gandhi, Singh, Swami, Ganesh, & Sinhaet, 2017; Grossmann, 2015; L. B. Smith, Jayaraman, Clerkin, & Yu, 2018; Turati, 2004; but see Young and Burton, 2018, for a more qualified claim). Faces offer a rich, salient source of information for navigating the social world: They play a role in deciding whom to love, whom to trust, whom to help, and who is found guilty of a crime (Todorov, 2017; Zebrowitz, 1997, 2017; Zhang, Chen, & Yang, 2018). 

Beginning with the ancient Greeks (Aristotle, in the 4th century BCE) and Romans (Cicero), various cultures have viewed the human face as a window on the mind. But to what extent can a raised eyebrow, a curled lip, or a narrowed eye reveal what someone is thinking or feeling, allowing a perceiver’s brain to guess what that someone will do next? The answers to these questions have major consequences for human outcomes as they unfold in the living room, the classroom, the courtroom, and even on the battlefield. They also powerfully shape the direction of research in a broad array of scientific fields, from basic neuroscience to psychiatry. 

Understanding what facial movements might reveal about a person’s emotions is made more urgent by the fact that many people believe they already know. Specific configurations of facial-muscle movements appear as if they summarily broadcast or display a person’s emotions, which is why they are routinely referred to as emotional expressions and facial expressions. A simple Google search for the phrase “emotional facial expressions” (see Box 1 in the Supplemental Material available online) reveals the ubiquity with which, at least in certain parts of the world, people believe that certain emotion categories are reliably signaled or revealed by certain facial-muscle movement configurations—a set of beliefs we refer to as the common view (also called the classical view; L. F. Barrett, 2017b). Likewise, many cultural products testify to the common view. Here are several examples:

  • Technology companies are investing tremendous resources to figure out how to objectively “read” emotions in people by detecting their presumed facial expressions, such as scowling faces, frowning faces, and smiling faces, in an automated fashion. Several companies claim to have already done it (e.g., Affectiva.com, 2018; Microsoft Azure, 2018). For example, Microsoft’s Emotion API promises to take video images of a person’s face to detect what that individual is feeling. Microsoft’s website states that its software “integrates emotion recognition, returning the confidence across a set of emotions . . . such as anger, contempt, disgust, fear, happiness, neutral, sadness, and surprise. These emotions are understood to be cross-culturally and universally communicated with particular facial expressions” (screen 3). 

  • Countless electronic messages are annotated with emojis or emoticons that are schematized versions of the proposed facial expressions for various emotion categories (Emojipedia.org, 2019). 

  • Putative emotional expressions are taught to preschool children by displaying scowling faces, frowning faces, smiling faces, and so on, in posters (e.g., use “feeling chart for children” in a Google image search), games (e.g., Miniland emotion games; Miniland Group, 2019), books (e.g., Cain, 2000; T. Parr, 2005), and episodes of Sesame Street (among many examples, see Morenoff, 2014; Pliskin, 2015; Valentine & Lehmann, 2015). 

  • Television shows (e.g., Lie to Me; Baum & Grazer, 2009), movies (e.g., Inside Out; Docter, Del Carmen, LeFauve, Cooley, and Lassetter, 2015), and documentaries (e.g., The Human Face, produced by the British Broadcasting Company; Cleese, Erskine, & Stewart, 2001) customarily depict certain facial configurations as universal expressions of emotions. 

  • Magazine and newspaper articles routinely feature stories in kind: facial configurations depicting a scowl are referred to as “expressions of anger,” facial configurations depicting a smile are referred to as “expressions of happiness,” facial configurations depicting a frown are referred to as “expressions of sadness,” and so on.  

  • Agents of the U.S. Federal Bureau of Investigation (FBI) and the Transportation Security Administration (TSA) were trained to detect emotions and other intentions using these facial configurations, with the goal of identifying and thwarting terrorists (R. Heilig, special agent with the FBI, personal communication, December 15, 2014; L. F. Barrett, 2017c). 

  • The facial configurations that supposedly diagnose emotional states also figure prominently in the diagnosis and treatment of psychiatric disorders. One of the most widely used tasks in autism research, the Reading the Mind in the Eyes Test, asks test takers to match photos of the upper (eye) region of a posed facial configuration with specific mental state words, including emotion words (Baron-Cohen, Wheelwright, Hill, Raste, & Plumb, 2001). Treatment plans for people living with autism and other brain disorders often include learning to recognize these facial configurations as emotional expressions (Baron-Cohen, Golan, Wheelwright, & Hill, 2004; Kouo & Egel, 2016). This training does not generalize well to real-world skills, however (Berggren et al., 2018; Kouo & Egel, 2016). 

  • “Reading” the emotions of a defendant — in the words of Supreme Court Justice Anthony Kennedy, to “know the heart and mind of the offender” (Riggins v. Nevada, 1992, p. 142) — is one pillar of a fair trial in the U.S. legal system and in many legal systems in the Western world. Legal actors such as jurors and judges routinely rely on facial movements to determine the guilt and remorse of a defendant (e.g., Bandes, 2014; Zebrowitz, 1997). For example, defendants who are perceived as untrustworthy receive harsher sentences than they otherwise would (J. P. Wilson & Rule, 2015, 2016), and such perceptions are more likely when a person appears to be angry (i.e., the person’s facial structure looks similar to the hypothesized facial expression of anger, which is a scowl; Todorov, 2017).

An incorrect inference about defendants’ emotional state can cost them their children, their freedom, or even their lives (for recent examples, see L. F. Barrett, 2017b, beginning on page 183). 

But can a person’s emotional state be reasonably inferred from that person’s facial movements? In this article, we offer a systematic review of the evidence, testing the common view that instances of an emotion category are signaled with a distinctive configuration of facial movements that has enough reliability and specificity to serve as a diagnostic marker of those instances. We focus our review on evidence pertaining to six emotion categories that have received the lion’s share of attention in scientific research—anger, disgust, fear, happiness, sadness, and surprise—and that, correspondingly, are the focus of the common view (as evidenced by our Google search, summarized in Box 1 in the Supplemental Material). Our conclusions apply, however, to all emotion categories that have thus far been scientifically studied. We open the article with a brief discussion of its scope, approach, and intended audience. We then summarize evidence on how people actually move their faces during episodes of emotion, referred to as studies of expression production, following which we examine evidence on which emotions are actually inferred from looking at facial movements, referred to as studies of emotion perception. We identify three key shortcomings in the scientific research that have contributed to a general misunderstanding about how emotions are expressed and perceived in facial movements and that limit the translation of this scientific evidence for other uses:

  • Limited reliability (i.e., instances of the same emotion category are neither reliably expressed through nor perceived from a common set of facial movements). 

  • Lack of specificity (i.e., there is no unique mapping between a configuration of facial movements and instances of an emotion category). 

  • Limited generalizability (i.e., the effects of context and culture have not been sufficiently documented and accounted for). 

We then discuss our conclusions, followed by proposals for consumers on how they might use the existing scientific literature. We also provide recommendations for future research on emotion production and perception with consumers of that research in mind. We have included additional detail on some topics of import or interest in the Supplemental Material.

Financial Surveillance

'Plastic surveillance: Payment cards and the history of transactional data, 1888 to present' by  Josh Lauer in (2020) Big Data and Society comments
Modern payment cards encompass a bewildering array of consumer technologies, from credit and debit cards to stored-value and loyalty cards. But what unites all of these financial media is their connection to recordkeeping systems. Each swipe sends data hurtling through invisible infrastructures to verify accounts, record purchase details, exchange funds, and update balances. With payment cards, banks and merchants have been able to amass vast archives of transactional data. This information is a valuable asset in itself. It can be used for in-house data analytics programs or sold as marketing intelligence to third parties. This research examines the development of payment cards in the United States from the late 19th century to present, drawing attention to their fundamental relationship to identification, recordkeeping, and data mining. The history of payment cards, I argue, is not just a history of financial innovation and computing; it is also a history of Big Data and consumer surveillance. This history, moreover, provides insight into the growth of transactional data and the datafication of money in the digital economy.
Lerner argues
In 2002, an executive at Canadian Tire, a Toronto-based national retailer, made some surprising discoveries about the store’s credit customers. Close analysis of the store’s card data revealed statistical correlations between how customers used their cards and how they paid their bills. Among other things, the executive learned that customers who purchased bird seed and protective felt pads for the feet of chairs tended to be reliable bill payers. By contrast, those who bought cheap motor oil and chrome skull auto accessories were more likely to miss payments. “If you show us what you buy,” the executive boasted to a New York Times reporter, “we can tell you who you are, maybe even better than you know yourself” (Duhigg, 2009). Several years later, marketing executives at Target, the American mass retailer, were making similar discoveries in a now infamous data-mining program. By linking the store’s guest identification numbers with purchase records, analysts were able to predict which shoppers were pregnant. Transactional data revealed correlations between pregnancy and a variety of purchases, including prenatal vitamin supplements, unscented lotion, and giant bags of cotton balls (Duhigg, 2012). 
When these stories broke, they illustrated the startling power of Big Data. While Target’s program attracted media attention and sparked privacy concerns, both cases showed how massive data sets could be mined for hidden behavioral clues and marketing insights. More importantly, they revealed how payment itself—point-of-sale (POS) swipes and card-based online sales—could be leveraged into consumer data programs. These retailers were not alone. During the first decade of the 21st century, payment systems were increasingly viewed as data goldmines. Unlike cash transactions, which produced only anonymous receipts, card payments yielded transactional data attached to specific individuals. This surplus data, merchants discovered, was its own reward. It could be used in-house, as with Canadian Tire and Target, or it could be reassembled, parsed, and sold as marketing intelligence to third parties. Payment cards were not just convenient stand-ins for cash; they were data-harvesting devices. 
Reflecting on recent developments in payment systems, anthropologist Bill Maurer (2014) posed a provocative question: “Is there money in credit?” Given the enormous analytical power of Big Data, credit’s informational yield has begun to compete with its value as a source of fees and interest. “Credit’s function,” Maurer mused, “may no longer be as money, but as a means to ever more consumer data” (513). Maurer’s observation points to a radical shift in the business logic of payment. Though money has always been linked to memory via recordkeeping, whether inscribed in tally sticks or account books (Maurer and Swartz, 2017), the records themselves have been subordinate to the value of the commodities, bills, or cash they represent. More recently, however, electronic payment systems have automated money’s recordkeeping function and vastly expanded the volume and granularity of transactional data that is recorded. As Rachel O’Dwyer (2019) argues, this is not simply an elaboration of money’s historical recordkeeping function; it is central to the emergent platform economy, which depends upon the corralling of user interactions so that their behaviors can be collected and monetized. “What is different here,” O’Dwyer writes, “is that … the monetary record becomes a proxy for the intimate secrets and desires of its users” (10). 
Building on Maurer and O’Dwyer’s insights, this article examines the long and complex evolution of the payment card’s surveillance function in the United States from the late 19th century to present. Modern payment cards encompass a bewildering array of consumer technologies, from credit and debit cards to stored value cards and loyalty cards. But what unites all of these financial media is their connection to recordkeeping systems. The simplicity of paying with plastic belies the complexity of information-processing infrastructures — the pipes and rails, in industry speak — that make it possible (Gießmann, 2018; Maurer, 2012). Each swipe, for a $4 coffee or a $40,000 wristwatch, sends data hurtling through invisible networks to instantly verify accounts, record purchase details, exchange funds, and update balances. Where cash facilitated “secrecy” and “concealment,” as Georg Simmel (1990: 385) once observed, payment cards demand the opposite: continuous exposure and confession. The history of payment cards, I argue, is not just a history of financial innovation and computing; it is also a history of consumer surveillance. 
This history, moreover, provides insight into the proliferation of transactional data in the modern surveillance economy. One of the defining characteristics of computerized information processing is the production of transactional data. The necessity of data “capture” is fundamental to the logic and design of modern computing (Agre, 1994). While electronic computers introduced new possibilities for capturing data, they also expanded the number and variety of data points generated during interactions themselves, including metadata (Schneier, 2015). This particular technological augmentation and related project of datafication (van Dijck, 2014) made it possible to compile more detailed records and ultimately to discover new uses for data, especially uses unrelated to the context of its original collection (Nissenbaum, 2010). The development of late 20th-century data mining and Big Data analysis would depend upon access to such accumulations of repurposed data (Kitchin, 2014; Mayer-Schönberger and Cukier, 2013). Transactional data would not only feed Big Data programs; it provided the building blocks of the modern surveillance economy. This account shows how data “capture,” a philosophical metaphor (Agre, 1994), has become indistinguishable from surveillance: the systematic, real-world collection of information for the purpose of social classification, prediction, and control. 
The discovery of transactional data and the rise of surveillance-based business models is often traced to contemporary tech giants. This perspective is exemplified by Zuboff’s (2015, 2019) account of “surveillance capitalism,” a concept that neatly captures the totalizing scale and exorable commercial logic of Big Data aggregators and platforms. For Zuboff (2019), the commodification of untapped transactional data—“behavioral surplus”—began in the early 2000s and can be credited to one company in particular: Google. “The discovery of behavioral surplus markets,” she writes, “marks a critical turning point not only in Google’s biography but also in the history of capitalism” (91). This narrative, however, is misleading. The value of transactional data was recognized much earlier, not by Google, but by other capitalists—namely, credit-granting department stores during the 1920s and credit card companies during the 1970s and 1980s. Both retailers and banks mined their payment records for insight into the buying habits, interests, and future profitability of their customers. The history of payment cards thus reveals the deep roots of surveillance capitalism and efforts to transform data into capital (Sadowski, 2019; West, 2017). 
By turning attention back on the payment card, this article examines the fundamental relationship between credit and transactional data, and the historical significance of payment cards as data-capturing technologies both before and after digitalization. Importantly, this account shows how payment systems, with their imperative to record all financial activities, provided a conceptual model for transactional data capture in the modern surveillance economy. Indeed, the inspiration for Google’s data-siphoning programs can be traced to electronic payment systems and their automatic generation of transactional data (Varian, 2010). After reviewing existing surveillance scholarship concerning credit and payment cards, I examine historical linkages between identification and recordkeeping in retail charge systems in the United States during the late 19th and early 20th centuries. I then describe technological changes in American payment card systems between the 1970s and 1990s, when electronic billing and POS systems enabled new forms of transactional data capture and suggested new modes of data monetization. Finally, I connect this history to the development of computerized transactional data, including Google’s pioneering data programs, and long-running privacy concerns surrounding electronic payment systems and their contribution to the intensification of consumer surveillance.