The iniquity exception to privilege: when instructing investigators costs you protection

What does the iniquity exception actually do?

The iniquity exception prevents legal professional privilege from attaching to communications made as part of, or in furtherance of, fraud or equivalent underhand conduct. Two 2026 decisions, Bourlakova v Anufriev [2026] EWHC 1747 (Ch) and Pliego v Astor Asset Management 3 Ltd [2026] EWCA Civ 940, now mark out where legitimate investigation ends and where instructing agents costs a litigant protection, or worse.

The doctrinal point matters because of how the exception operates. It does not strip privilege from documents that once had it; as Lewison LJ explained in Addlesee v Dentons Europe LLP [2019] EWCA Civ 1600 at [29], it prevents privilege from arising in the first place, so a client knows at the moment of communicating whether the protection exists. Richards J adopted that analysis in Bourlakova at [17], and it frames everything that follows: the question is never whether iniquity justifies disclosure of privileged material, because material caught by the exception was never privileged at all.

Who this guide is for

This explainer is written for civil fraud and asset recovery practitioners, disputes lawyers instructing enquiry agents or considering whether an opponent’s instruction of them opens a line of attack, and compliance officers setting a firm’s policy on investigators. It assumes familiarity with the two limbs of legal professional privilege and concentrates on the 2026 authorities, read against Al Sadeq v Dechert LLP [2024] EWCA Civ 28, which remains the governing statement of the exception.

When does iniquity defeat privilege?

Richards J set out three ingredients at [18] of Bourlakova, each of which must be present. There must be documents or communications that would otherwise attract privilege; an iniquity established to the requisite standard; and the requisite connection between the iniquity and the documents. On connection, Al Sadeq at [166] holds that where a prima facie case of iniquity engages the exception, there is no privilege in documents brought into existence as part of or in furtherance of it, which includes documents preparing for or reporting on the iniquitous conduct, though not documents that merely would not exist but for it.

Iniquity itself does not require fraud or dishonesty. Popplewell LJ put it at [55] of Al Sadeq as fraud or other equivalent underhand conduct in breach of a duty of good faith or contrary to public policy or the interests of justice, drawing on Barclays Bank plc v Eustice [1995] 1 WLR 1238, where transactions at an undervalue designed to prejudice a creditor under section 423 of the Insolvency Act 1986 sufficed without any finding of dishonesty. Some disreputable conduct still falls short: inducing a breach of contract did not qualify in Crescent Farm, and trespass to chattels did not in Dubai Aluminium Co Ltd v Al Alawi [1999] 1 WLR 1964.

Iniquity alone is still not enough where the communications involve lawyers. The touchstone from [58] of Al Sadeq, applied in Bourlakova at [28] to [29], is whether the iniquity takes the relationship outside the normal scope of a professional engagement or abuses a relationship within it. A client instructing solicitors to advance a case the client knows to be false remains, remarkably enough, within the ordinary run, which is why the exception is narrower than instinct suggests.

The standard of proof at an interlocutory stage is the balance of probabilities on the material available, however imperfect that exercise is without cross-examination ([36] to [40] of Bourlakova, following Al Sadeq at [108]). Richards J was prepared to make findings about a transcript that neither he nor the claimants had seen, on the strength of independent review counsel’s assessment, precisely because refusing to make any finding would have forced the applicants to give away the privilege they were asserting ([50] to [51]).

Do your beliefs about the investigation matter?

Yes, and this is the practical heart of Bourlakova. The applicants argued that Eustice made a client’s belief in the lawfulness of the conduct irrelevant. Richards J rejected that reading at [30] to [35], holding that beliefs are relevant in two respects: they bear on whether there is iniquity at all, since underhand conduct in breach of good faith invites consideration of mental state, and they bear on whether the solicitor and client relationship stayed within the ordinary run.

On the facts, that analysis carried the day. The Bourlakovas and their solicitors had instructed a reputable investigation firm on contractual terms that required lawful conduct, including an express obligation to comply with Imerman v Tchenguiz [2010] EWCA Civ 908 ([60] to [61]); they believed material was coming from a whistleblower rather than from hacking ([133] to [139]); and when they discovered in August 2023 that the agents had obtained potentially privileged material, they instructed independent review counsel so that neither client nor solicitors would see it ([72] to [76]). The judge declined to find any of the four alleged iniquities established ([196]), and refused to recognise a rule that it is iniquitous per se to put in train a process capable of obtaining confidential information without consent ([167]), because such a rule would risk a chilling effect on the ability of claimants to gather evidence of suspected fraud ([168]).

Two cautions temper the comfort. The disclosure figures show how much turns on these findings: of 2,039 potentially relevant documents identified in the agents’ disclosure exercise, only 178 were disclosed, with privilege asserted over the remaining 1,861 ([84]). And the eavesdropping instruction to a second firm of agents, who listened to the opposing parties’ meeting with their own lawyer in a restaurant, was described at [194] as towards the margin, surviving only because iniquity requires more than conduct that is disreputable or a failure to maintain good ethical standards. A differently minded judge, or slightly worse facts, could move that outcome.

What separates investigation from privilege hunting?

Eleven days after Bourlakova, the Court of Appeal answered the harder question in Pliego. There, the claimants engaged Black Cube under a retainer of £1,144,000 plus a 4% success fee to run a covert sting against the partner conducting the defendants’ case, whose operative posed as a prospective client and, across three recorded meetings totalling around six and a half hours, drew the solicitor into revealing litigation strategy, settlement strategy and the pressure the freezing orders were exerting.

Males LJ, with whom Phillips and Foxton LJJ agreed, had no doubt that this was a very serious abuse of process, describing the suborning of an opponent’s solicitor by deception as a form of corruption at [79] and observing that some things are so obvious that they do not need much analysis ([74]). It made no difference that the solicitor was himself at fault, or that some of the information might ultimately have fallen within the iniquity exception, because the claimants could not have known that when they sanctioned the operation ([75] to [77]). The court also doubted the assumption below that nothing unlawful had occurred, noting at [78] that the conduct may well have amounted to fraud by false representation under section 2 of the Fraud Act 2006.

The response is where Pliego will be cited for years. The Court of Appeal refused to make strike-out the default answer to privilege hunting, reasoning at [91] to [92] that other equally culpable abuses, suppressed disclosure, perjured evidence, unlawful hacking, do not attract one, and that a default rule would turn the last resort into the first. Instead it discharged every freezing order the claimants had obtained, because the orders had been continued while the claimants stayed silent about an operation the judge would never have countenanced ([99] to [106]), and it left the fraud claim to go to trial. The postscript at [112] warns solicitors presented with the fruits of such an operation that using the material may encourage, or assist a client to benefit from, the commission of criminal offences.

Read together, the two judgments draw a workable boundary. A litigant who instructs agents on lawful terms, holds a documented and reasonable belief in the lawfulness of their methods, and builds a mechanism to keep privileged material away from the legal team sits inside the line, even where the investigation in fact yields an opponent’s confidential material. A litigant who targets the opponent’s lawyer, or who sets out to acquire the privileged thinking on the other side, sits well outside it, and the exposure is not limited to losing privilege: the court will strip discretionary remedies, the trial judge may deny equitable relief, and the criminal law is not safely out of reach.

Conduct compared: inside and outside the line

A checklist before instructing investigators

• Put lawfulness in the contract, expressly and in your own terms, including compliance with Imerman and the civil and criminal law of every jurisdiction the investigation will touch, as the solicitors did in Bourlakova at [60].
• Record, at the point of instruction, why the client believes wrongdoing has occurred and why the firm considers the proposed methods lawful, because the court will examine beliefs and a contemporaneous note is worth more than a witness statement written after challenge.
• Define the remit around evidence of the suspected wrongdoing, and never instruct or permit agents to seek the opponent’s privileged material or to approach the opponent’s lawyers on any pretext.
• Agree in advance what happens if the agents come into possession of potentially privileged material: quarantine, no transmission to the legal team, and review by counsel independent of the litigation, the mechanism Richards J found unobjectionable at [180] to [191].
• Take a decision, with reasons recorded, on when the opposing party must be notified that their material has been obtained, recognising that deferral was defensible in Bourlakova at [182] but that silence while holding interim relief proved fatal to the freezing orders in Pliego at [105].
• Revisit the instruction whenever the picture changes, because the finding that the claimants acted properly in August 2023 rested on the steps they took the moment they learned what their agents held.
• If offered the fruits of someone else’s operation, stop and take the [112] postscript in Pliego seriously before reviewing anything, since the reviewing solicitor’s belief that privilege had been waived was described by Males LJ as obviously untenable.

Common mistakes

• Assuming that because the agents obtained confidential material, privilege in the instruction communications is lost. Bourlakova at [167] rejects exactly that per se rule.
• Treating a contractual promise of lawfulness as the end of the exercise rather than the beginning, when the judgment weighed the whole course of conduct, including how the firm reacted on discovery.
• Relying on the argument that the material would have been disclosable anyway. That reasoning helped in the hacking case the Court of Appeal distinguished, but Males LJ held at [98] that it has no purchase where what was taken is the opponent’s litigation strategy.
• Believing that the other side’s solicitor being at fault dilutes the abuse. The court in Pliego at [75] called that submission absurd, since the solicitor’s indiscretion was the whole object of the operation.
• Forgetting that outcome and reasoning diverge: both 2026 cases involve investigators and privilege, yet one claimant kept every protection and the other lost worldwide freezing orders, on facts that a briefing note could easily blur together.

Official sources

The full judgments are on Find Case Law: Bourlakova v Anufriev [2026] EWHC 1747 (Ch) and Pliego v Astor Asset Management 3 Ltd [2026] EWCA Civ 940, alongside the governing Court of Appeal statement in Al Sadeq v Dechert LLP [2024] EWCA Civ 28. Our news coverage of the first instance decision is at Privilege survives the enquiry agent: Bourlakova and the iniquity exception, and the disclosure context sits alongside our guide to pre-action disclosure under CPR 31.16.

Date last updated: 10 August 2026.