High Court decisions show developing court approach to dishonest assistance and unjust enrichment claims

The two cases involved e-money institution Revolut - Larsson v Revolut and Terna Energy v Revolut (7 pages / 188 KB).

The first essential element of dishonest assistance is that there must be a breach of trust. In a fraud context, the analysis is that when the victim’s money comes into the fraudster’s hands the fraudster does not have legal ownership of the money. Instead, the fraudster holds it as a constructive trustee and is subject to a duty in law to return it to the victim of the fraud.

The second essential element is that there has to be a breach of trust. If the fraudster pays the money away then the analysis would be that there is a dealing with trust property, and that dealing is in breach of the trust because the only permissible dealing would be to return the money to the victim of the fraud.

By characterising the proceeds of the fraud as trust property the law is able to trace the proceeds into the hands of the next person who receives the money – so long as that person is aware of the fraudulent origin of the money. If a third party assists the fraudster to commit the fraud then the third party becomes equally liable as the fraudster, but only if the third party has sufficient knowledge of the fraud for the assistance to be classed as dishonest.

In the first case, Larsson brought a claim against Revolut alleging that Revolut breached its contractual duties by failing to detect and prevent fraud. Larsson made payments from his Swiss UBS account to five accounts that fraudsters had set up at Revolut. The purported objective for the transactions was to purchase shares in a company, with the fraudsters telling Larsson his money would not be at risk because it would be held in accounts in his own name until it was time to transfer the money out to buy the shares.

The money transfers were international, so they were made using account numbers sent by SWIFT message as opposed to the account number plus account holder name system that is used for domestic transfers. The SWIFT messages said that the payments were for the benefit of Larsson, but the accounts were actually in the names of the fraudsters. Revolut credited the payments to the accounts before spotting the issue and stopping the final payment.

Larsson argued that Revolut dishonestly assisted in a breach of trust by crediting the payments to the accounts in the fraudsters’ names despite the SWIFT messages saying that the payments were for the benefit of Mr Larsson. The judge held that there was no contractual duty on Revolut to check the names on the accounts because this is not how international payments initiated by SWIFT messages work. However, the judge did not think that the case was clear enough to strike out the dishonest assistance part of the claim. The judge held that it was too early to say whether on a full investigation there might be some evidence suggesting that Revolut knew about the fraud or turned a blind eye to it sufficiently to meet the dishonesty test. The case will have to continue to trial and be decided on the facts.

Caroline Hearn, banking litigation expert at Pinsent Masons said: “The dishonest assistance argument is not a new law. Banks have always been at risk of liability if they transfer or receive money that they know or have good reason to suspect is the proceeds of fraud and the bank’s conduct meets the dishonesty threshold. However, it is difficult to win on this argument as usually the bank does not know anything at all about the underlying fraud, so it is impossible to prove it had dishonest knowledge.”

Mike Hawthorne, banking disputes expert at Pinsent Masons, said: “The decision shows that, in theory, a bank can become liable for dishonest assistance where it allows it customers to receive and pay proceeds of crime in circumstances where it knew or should have known of the crime. The bigger picture, however, is that this appears to be an allegation of dishonesty based on what appears to what was probably at worst a process failure. There was no evidence that Revolut knew anything about the underlying fraud, but the allegation alone was enough to allow the dishonest assistance part of the claim to progress towards trial.”

The second case, Terna Energy v Revolut, involved a claim for unjust enrichment to recover €700,000. The fraudster tricked Terna into paying cash to a fake supplier, with the payments processed through Revolut. The payment was initially frozen but later released by Revolut, with Terna claiming Revolut took no steps to investigate the reasons for the payment.

The essential elements of unjust enrichment are that the receiving party has to be enriched at the expense of the receiving party, and it must be unjust for the receiving party to keep hold of the money. If the paying party can prove those elements then there is in principle a liability on the receiving party to repay the money even if the receiving party actually paid the money away to a third party.

Unjust enrichment was previously considered an almost impossible claim to make against a bank unless there were very unusual circumstances. The basis rationale was that, especially with international payments, the receiving bank is usually at the end of a long chain of transfers involving other banks, so it is hard to say that it has been “enriched” at the expense of the person who originated the payment, as opposed to being enriched at the expense of the last but one bank in the chain.

Even if the claimant could get over that problem, the next line of defence for the receiving bank would be that it was not enriched by the incoming payment at all, because it only received the money on behalf of its customer, and as soon as the money comes into the account the bank has a matching liability to pay out to its customer.

However, the judge in Terna disagreed with the conventional analysis and held that it was at least possible for a bank, in this case Revolut, to be enriched by an incoming payment at the expense of the defrauded party. As regards the chain of payments point, the judge held that the real nature of the transaction is a payment from A to B, and the fact that A and B used agents (the intermediate banks) to facilitate the transaction is a matter of process only As regards the matching liability point, the judge held that if the receiving bank knew that the payment had been procured by fraud then the bank would not have an obligation to account to its customer for the payment if it knew that the payment had been produced by fraud, so in that scenario there would be no legal liability between the bank and the customer, and the bank would itself be “enriched” by the incoming payment.

Revolut has been granted permission to appeal the Terna Energy case to the Court of Appeal. The judge recognised that his analysis was opposed to previous analyses by judges at the same level and said, “Given the vast numbers of transfers of funds which take place through the banking system every day, and the need for clarity as to their legal effect, these are issues which will arise again and again, and must sooner or later be resolved at appellate level”.

“Obviously none of this comes into play until the receiving bank has some reason to suspect fraud, but that just begs the question of what knowledge would be enough to create suspicion of fraud. Pending the appeal this decision creates uncertainty for receiving banks about what potential liabilities could flow from their fraud detection systems flagging indicators of fraud on an incoming payment. There is also a question mark over how much both paying and receiving banks have to do in order to comply with the still unknown scope of the ‘retrieval duty’, for example taking steps to stop the money by notifying the next bank down the line of the suspected fraud. We can expect paying banks to err on the side of caution when they are told about fraud by their customer and pass on the information through to the receiving bank even faster than they did before. Receiving banks could then be at risk of unjust enrichment claims if they allow their customers to use the money before resolving the fraud markers,” said Hawthorne.

“These two cases area also a useful reminder of the distinction between a summary judgment and striking out,” said Nicola Seymour, litigation expert at Pinsent Masons. Strike out applications are generally determined by reference to the facts in the statements of case, rather than the determination of any disputed matters of fact as can be done in a summary judgment application.

In the Larsson case, while certain pleading defects regarding the dishonest assistance claim were noted, the strike out application ultimately failed as the Judge was not prepared to find that there was no sufficiently arguable case.  “A much more detailed argument was required in order to resolve the question, which could have a wider impact on other cases,” said Seymour.

 “Strike out and summary judgement applications can be risky when you are dealing with something that has systemic importance. You can end up with interim decisions that analyse the law without applying it to the real facts, or that conclude that the law is too difficult to resolve on a strike out. Either way the consequence of a failed strike out application is often uncertainty.  When a contested legal issue could have ramifications for you beyond the case in question it’s often wiser to run a case to trial rather than to risk a strike out application, especially if you think the underlying facts will be favourable when everything comes out at trial,” said Hawthorne.