THE test of dishonesty and the role of knowledge in dishonest assistance has been considered by the Court of Appeal in Group Seven Ltd. v Notable Services L.L.P. [2019] EWCA Civ 614, [2019] 3 W.L.R. 1011. Allseas Group S.A. is a Swiss undersea pipe-laying company. Allseas had a brilliant idea: it would build a ship that could sail to an oil rig, lift it out of the water, and repair it. A group of fraudsters informed Allseas of a secret investment scheme with superlative returns supposedly run by the US Federal Reserve. It was only available to special investors with investment projects that were exceptionally beneficial to humankind. Fortunately, Allseas was special: the Federal Reserve wanted to help Allseas with its ship. Allseas decided to invest €100 m in the scheme through a subsidiary, Group Seven Ltd. Mr. Louis Nobre was one of the fraudsters. He created a company, Larn Ltd., to receive the €100 m as a “loan” from Group Seven. Having obtained control of the €100 m, Mr. Nobre then attempted to launder it through the respectable-looking client account of Notable Services L.L.P., a London law and accounting firm. He needed to persuade Notable to take him on as a client without revealing too much about the source of the money. So he enlisted the aid of the relationship manager of a small Swiss Bank to provide false references to Notable. The banker was paid a bribe for his services. Mr. Nobre also persuaded an accountant at Notable to help him gain Notable's confidence. The accountant was paid a bribe of £170,000 out of the client account. It was styled a “personal fee” and paid to a Panamanian company the accountant controlled. Rule 14.5 of the Solicitor's Account Rules 2011 prohibits the provision of banking facilities through a client account. Notable took external legal advice as to whether the firm could proceed with the transactions, and was advised that on the assumed facts (including the references by the banker) they could. The accountant concealed the existence of the bribe from the firm and the firm's legal advisers. The advice would undoubtedly have been different if the legal advisers had known about this bribe.

This loan was declared a sham and rescinded in earlier litigation: [2014] EWHC 2046 (Ch). Upon rescission, the beneficial title to the money retrospectively revested in the claimant under a constructive trust: National Crime Agency v Robb [2015] Ch. 520, at [49]. All dealings with the money by Larn and Mr. Nobre were therefore a breach of trust and the Court of Appeal treats this constructive trust arising from rescission as a sufficient foundation for claims of dishonest assistance.

The trial judge held the accountant liable for knowing receipt of the £170,000 bribe on the ground that a reasonable person in the position of the accountant would not have thought that Mr. Nobre was entitled to the €100 million. But when it came to dishonest assistance, the trial judge held that the accountant did not, at the relevant time, have “blind eye knowledge” that Mr. Nobre was not entitled to use the €100 million, and thus that the accountant was not dishonest. The phrase “blind eye knowledge” was taken from Manifest Shipping & Co. Ltd. v Uni-Polaris Insurance Co. Ltd. [2001] UKHL 1, [2003] 1 A.C. 469. Group Seven appealed to the Court of Appeal, which held that the trial judge had approached the question of dishonesty in too compartmentalised a fashion, and had left out of the overall evaluation the fact that the accountant had received a bribe and did not tell the other members of the firm even when external legal advice was being sought. When that was taken into account, the accountant did have “blind eye knowledge” of the breach and was therefore dishonest.

Lord Nicholls’ judgment in Royal Brunei Airlines Sdn Bhd v Tan [1995] 2 A.C. 378 is usually taken to have recast the fault element of accessory liability, replacing the overly-refined and artificial tests of “knowledge” from Baden, Delvaux & Lecuit v Société Générale pour Favoriser le Développement du Commerce et de l'Industrie en France S.A. [1993] 1 W.L.R. 509, at [250] with a simpler test of “dishonesty”. Knowledge was best left out as an element of liability, and Baden was best “forgotten”. The shift in the intellectual climate was reflected in the textbooks which duly retitled the liability from “knowing assistance” to “dishonest assistance”. But in the haste to make a clean break from the old authorities on knowledge, there was little authority from which Lord Nicholls could draw to define the new standard of dishonesty, and so the discussion of dishonesty in Tan was done in global terms, without much particularisation to the facts of a modern dishonest assistance case.

Discussion of the concept of “dishonesty” should not be divorced from the other elements of assistant liability. When the term is used in the context of dishonest assistance, it is used in an adjectival or adverbial form. So the sort of dishonesty that the court is concerned to find is the sort that must be involved when assistance in a breach of trust or fiduciary duty is already established. It is not an enquiry into dishonesty simpliciter. A similar point has been made as to the standard of fault in the doctrine of unclean hands, where the rule is that “a man must come into a Court of Equity with clean hands; but when this is said, it does not mean a general depravity; it must have an immediate and necessary relation to the equity sued for” (Dering v Earl of Winchelsea (1787) 1 Cox. 318, 319, 29 E.R. 1184, 1185).

When a person has assisted in a breach of fiduciary duty, the factual variable that is involved in making an assessment of dishonesty is the assistant's level of knowledge of the breach of trust or fiduciary duty. It is difficult to see how it would be possible to conclude that assistance was “dishonest” unless there was some finding about knowledge of the breach. Once that is realised, Tan can be seen as less of a revolution and more of a change in focus. The previous authorities, exemplified in the Baden scale, had drawn fine distinctions between different levels of “knowledge” and “notice” but were in a sense blind as to how to fix an appropriate level to the liability. The standard of dishonesty is similarly useless unless it is formulated with respect to the appropriate factual context, which is the state of the assistant's knowledge of the breach of trust or fiduciary duty. It is therefore unsurprising to see discussions of “blind eye knowledge” returning to the case law, but only as a preliminary to determining whether the assistant was dishonest.

The Court of Appeal in Group Seven, following dicta in Ivey v Genting Casinos UK Ltd. [2017] UKHL 67, [2018] A.C. 391, at [74], held that dishonesty involves a two-step test: the tribunal of fact must first make factual findings as to the assistant's state of knowledge, and then decide whether an honest and reasonable person would think the assistant's conduct was dishonest in light of that state of mind. The Court of Appeal in Group Seven confirms that the second element in this test is to be treated as settled by the Supreme Court in Ivey to be an objective test.

The first step in this test is now likely to be the main focus for future cases on dishonest assistance. In fraud cases it is rarely possible to discover explicit evidence as to the state of mind of the parties. In Edgington v Fitzmaurice (1885) 29 Ch.D. 459, 483, Bowen L.J. famously declared that “the state of a man's mind is as much a fact as the state of his digestion”. Maybe so. But it is easier to read a stomach than a mind. What is dishonest is seldom made explicit, for dishonest people are usually conscious of the need to hide their behaviour, or to cloak it in the raiment of equivocation. The explicit evidence as to knowledge at any particular time may appear quite meagre. The true state of the assistant's mind may only become apparent as an inference from the conduct of the assistant, viewed in total, forming a pattern over time, in light of the likely motives of the parties: Mortgage Agency Services Number One Ltd. v Cripps Harries L.L.P. [2016] EWHC 2483 (Ch), at [88]. So in making the factual findings for the first step of the test in Ivey, a court may need to thread the individual findings on the assistant's state of mind together by making a single broad finding of fact about the assistant's knowledge of the breach. Group Seven shows that a finding of “blind eye knowledge” will indicate dishonesty. Since dishonesty is an objective test, it is also arguable that when an assistant knows facts from which an honest person would have inferred that there was a breach of trust or fiduciary duty, but the fiduciary failed to draw that inference, a finding of dishonesty might also be made.