Dishonest assistance

Dishonest assistance is the principal equitable wrong by which a stranger to the trust is made personally liable for assisting a breach of fiduciary duty. It is not a proprietary claim, and it is not a species of tracing. It matters precisely where tracing and knowing receipt, covered in Week 14, run out. A dishonest assistant may never have received trust property; he may have provided banking facilities, drafted documents, procured companies, reassured victims, or stood behind an offshore structure. Equity’s response is personal: the assistant is liable to compensate for loss caused by the breach which he dishonestly assisted, and in appropriate cases may be ordered to account for profits made through the assistance.

The modern cause of action is usually traced to the second limb of Barnes v Addy (1874) LR 9 Ch App 244. The old language is that strangers may be liable if they assist with knowledge in a dishonest and fraudulent design on the part of the trustees. Modern law has moved away from the artificial search for a trustee’s dishonest design. Royal Brunei Airlines Sdn Bhd v Tan [1995] 2 AC 378 made the assistant’s dishonesty the touchstone. Barlow Clowes International Ltd v Eurotrust International Ltd [2005] UKPC 37 and Ivey v Genting Casinos (UK) Ltd [2017] UKSC 67 then clarified the test for dishonesty: establish the defendant’s actual knowledge or belief as to the facts; then ask whether, in the light of that state of mind, the conduct was dishonest by the standards of ordinary decent people. There is no separate requirement that the defendant appreciated that ordinary people would call his conduct dishonest.

For Cambridge Part II purposes the topic sits at the junction of three themes. First, the trust is enforced not only against trustees but also against outsiders whose participation threatens the integrity of equitable administration. Secondly, personal liability in equity is not unified: dishonest assistance must be distinguished from knowing receipt, inducement of breach of contract, deceit, conspiracy, unjust enrichment, and accessory liability in tort. Thirdly, the modern law illustrates equity’s oscillation between conscience and deterrence. It invokes conscience through dishonesty, but its remedial consequences may be strict and extensive once that threshold is crossed.

In an examination answer, the strongest structure is: identify the primary wrong; identify assistance; apply the dishonesty test carefully to actual facts known or believed by the defendant; consider causation, scope of liability, and remedy. Do not begin and end with moral adjectives. Dishonesty is not carelessness, nor commercial sharpness, nor a mere failure to investigate. But deliberate closing of eyes, suspicious participation, contrived ignorance, and knowingly facilitating a misapplication of assets are paradigmatic cases. Good Cambridge answers show why the defendant’s role is normatively serious enough to justify accessory liability, without converting every negligent professional or banker into a constructive trustee.

The starting point is Barnes v Addy. Lord Selborne LC insisted that persons dealing with trustees as agents or professional advisers are not to be made constructive trustees merely because they act in transactions within the trustees’ powers. Equity does not impose trusteeship on every solicitor, banker, broker, or agent who becomes involved in trust dealings. But Lord Selborne recognised two exceptional categories: receipt of trust property with sufficient knowledge, and assistance in a dishonest and fraudulent design. Those two categories became known as the two limbs of Barnes v Addy: knowing receipt and knowing assistance. The terminology has since altered. The first limb is now treated as a receipt-based personal claim requiring beneficial receipt and unconscionability. The second is dishonest assistance.

The historical language was awkward in three respects. First, it spoke of the stranger as a constructive trustee. That label is misleading. The defendant is not usually entrusted with property and does not assume the office of trustee. He is personally liable as an accessory to breach of fiduciary duty. Secondly, the old formula required assistance in a dishonest and fraudulent design of the trustee. That suggested that the primary wrongdoer had to be dishonest, even though a trustee may commit a breach innocently or negligently. Thirdly, equity experimented with classifications of knowledge, especially in the twentieth century, in an attempt to decide when participation was sufficiently culpable.

Selangor United Rubber Estates Ltd v Cradock (No 3) [1968] 1 WLR 1555 represents the expansive period. Megarry J treated constructive notice and knowledge of circumstances which would put an honest and reasonable person on inquiry as sufficient in the context of a fraudulent acquisition of corporate assets. Baden, Delvaux and Lecuit v Société Générale pour Favoriser le Développement du Commerce et de l’Industrie en France SA [1993] 1 WLR 509 famously classified knowledge into five categories: actual knowledge, wilfully shutting one’s eyes, wilfully and recklessly failing to make inquiries, knowledge of circumstances indicating the facts to an honest and reasonable person, and knowledge of circumstances putting an honest and reasonable person on inquiry. That Baden taxonomy was influential but unstable. It encouraged mechanical pleading: claimants alleged category 1 to 5 knowledge, and courts tried to decide where the facts fitted.

Royal Brunei reoriented the law. Lord Nicholls rejected the Baden categories as the organising principle for assistance. The core question is dishonesty on the part of the assistant. Knowledge remains important, but as evidence and as part of the factual matrix against which honesty is assessed. The claimant need not prove that the trustee was dishonest; it is enough that there was a breach of trust and that the defendant dishonestly assisted it. This severed accessory liability from the moral condition of the trustee and located the wrong in the assistant’s own conscience.

Twinsectra Ltd v Yardley [2002] UKHL 12 appeared to reintroduce a subjective element by suggesting that the defendant must realise that what he was doing would be regarded as dishonest by ordinary standards. Barlow Clowes then confined that reading, and Ivey has now settled the general test of dishonesty across civil and criminal contexts. The modern history is therefore one of movement from constructive trusteeship and notice, through taxonomies of knowledge, to a single dishonesty standard applied to the assistant’s actual knowledge or belief about the facts.

The elements of dishonest assistance are commonly stated as four: there must be a trust or fiduciary relationship; there must be a breach of trust or fiduciary duty; the defendant must have assisted that breach; and the assistance must have been dishonest. Each element repays close analysis.

First, the primary wrong need not be a breach by an express trustee. The cases include company directors, agents, solicitors, and others owing fiduciary duties. The language of trust remains central because the jurisdiction grew from trust administration, but the modern claim is not confined to technical trusts. A breach of fiduciary duty suffices. Conversely, one must identify a real primary wrong. Dishonest assistance is accessory liability; it is not a free-standing equitable wrong consisting merely of reprehensible conduct. A supervision essay should state the duty, the person who owed it, the act constituting breach, and the way in which the defendant’s conduct related to it.

Secondly, the trustee or fiduciary need not have been dishonest. This is a crucial consequence of Royal Brunei. If a trustee innocently misapplies trust funds and an outsider dishonestly facilitates that misapplication, the outsider may be liable. The wrong is the assistant’s dishonest participation in a breach of duty, not complicity in the trustee’s state of mind. That said, in many problem questions the primary breach will itself be fraudulent. Do not waste time proving dishonesty twice unless the facts require it.

Thirdly, assistance must be more than mere background presence. It need not be the sole cause of the breach, nor must it be indispensable, but it must be assistance in fact: conduct which helps, facilitates, procures, enables, conceals, or completes the breach. Examples include arranging transfers, providing bank accounts, drafting sham documents, acting as nominee, making false confirmations, moving money through corporate vehicles, or suppressing information. The requirement is not usually difficult where the defendant’s acts form part of the transaction by which trust assets are misapplied. It becomes difficult where the defendant’s role is remote: ordinary banking operations, routine professional services, or passive silence. The proper inquiry is whether the conduct materially advanced the breach or its implementation.

The doctrinal spine begins with Barnes v Addy. Its enduring significance is not the label constructive trustee, but the insistence on a limiting principle for strangers. Equity must protect beneficiaries from collusive outsiders without making routine advisers insurers of trust administration. The two limbs of Barnes v Addy supplied that limiting structure, though modern law has refined both limbs substantially.

Selangor United Rubber Estates Ltd v Cradock (No 3) illustrates the older, notice-based approach. In a corporate asset-stripping scheme, Megarry J was prepared to impose liability on a bank and other participants on broad knowledge principles. The case is historically important because it shows equity responding to commercial fraud by expanding constructive trusteeship. It is less reliable as a modern statement of the dishonesty test after Royal Brunei and Barlow Clowes.

Royal Brunei is the decisive modern authority. A travel agent held airline ticket proceeds on trust for the airline but paid them into its overdrawn account. The defendant company director controlled the agent and procured the use of the funds. The Privy Council held him liable. Lord Nicholls rejected the idea that the trustee’s dishonest and fraudulent design was an essential ingredient. The assistant’s own dishonesty is sufficient. The case also disapproved treating negligence or constructive notice as enough; dishonesty is the necessary threshold.

Twinsectra produced temporary uncertainty. A solicitor gave an undertaking that loan monies would be used only for acquiring property, but the funds were released for another purpose. The House of Lords held, by majority, that the defendant solicitor was not dishonest. Lord Hutton’s formulation appeared to require both objective dishonesty and the defendant’s realisation that his conduct was dishonest by those standards. That formulation was criticised because it allowed a morally obtuse defendant to escape liability.

Barlow Clowes restored the objective core. The Privy Council treated Twinsectra as not requiring a defendant to appreciate that his conduct was dishonest according to ordinary standards. The defendant’s knowledge or belief about facts is subjective; the standard of honesty is objective. Starglade Properties Ltd v Nash later applied this understanding in domestic civil litigation.

Ivey, though not a trusts case, now supplies the general test. The Supreme Court rejected the second limb of the criminal test in R v Ghosh and articulated the two-stage approach now used across civil law. Its importance for Trusts is that it confirms Barlow Clowes and removes the Twinsectra ambiguity.

Novoship and Group Seven are important for remedy and contemporary application. Novoship confirms that a dishonest assistant may be required to disgorge profits, but only profits sufficiently connected with the wrongdoing. Group Seven shows how modern courts analyse dishonest assistance in sophisticated frauds involving escrow arrangements, sham transactions, and professional intermediaries. These cases are especially valuable in Part II essays because they prevent the topic from becoming a purely historical discussion of Barnes, Royal Brunei, and Twinsectra.

See practice questions section below.

See further reading section below.