Pay Now, Prove Later: D&O Defence Costs After Liberty v Chedid

In Liberty Managing Agency Ltd & Ors v Chedid & Anor [2026] EWHC 2354 (Comm), the Commercial Court ordered D&O insurers on the second excess layer (which provided cover of £45m in excess of £30m) to advance defence costs to  two former Petrofac officers notwithstanding unresolved allegations of placement fraud.

The trial addressed three key issues: (1) the effect of non-avoidance wording; (2) the allocation of risk pending determination of alleged fraud; and (3) the insurability of costs incurred in defending criminal proceedings.

The decision is subject to a potential appeal, with permission to appeal having been granted.

Background

The Claimants were second excess layer D&O insurers of Petrofac Ltd and its directors and officers. The Defendants, Mr Marwan Chedid and Mr George Salibi, are former Petrofac officers facing bribery charges under the Bribery Act 2010, with trial listed at Southwark Crown Court from 2 November 2026.

With the primary and first excess layers close to exhaustion following the advancement of £30m in defence costs, the Insurers of the second excess layer sought to avoid the policy against the Defendants, alleging fraudulent misrepresentation and non-disclosure at placement – allegations which are disputed and, as yet, unproven. The Defendants said they could not fund their own defence without the second excess layer, and applied for an expedited trial of three preliminary issues:

• whether the Insurers were entitled to avoid the policy under Clause 8.2;
• whether the defence costs were uninsurable given the bribery allegations; and
• whether the Insurers were obliged to keep advancing costs pending determination of the fraud and bribery allegations. Mr Justice Jacobs found for the Defendants on all three issues.

Non-avoidance wording determines when avoidance rights may be exercised

The key question was the purpose of Clause 8.2. It said the insurer “shall not avoid this policy” unless the insured’s fraud “is established by a final decision of a court, tribunal or regulator or by a formal written admission“. The Insurers said this added nothing to the ordinary position: they could still avoid immediately and justify it later. The Court disagreed.

Ordinarily, avoidance for fraud is a ‘self-help’ remedy: the insurer avoids first and takes its chances that the avoidance is sustainable. The Court held that Clause 8.2 changed this because the words “is established by” imposed a timing condition which meant the insurer could not avoid until fraud had been established by one of the specified routes. The conduct exclusion in Clause 5.1 used the same formula, and the Court read both as part of a single scheme which had the effect that allegations are tolerated for the time being with any consequences following only after adjudication.

This is a reminder that non-avoidance clauses and conduct exclusions need to be read together. Depending on their wording, they may create a genuine condition precedent that prevents avoidance until fraud is independently proven, or they may simply preserve the insurer’s ordinary right to avoid, subject to later challenge.

Public policy permits parties to allocate the risk of unproven fraud

Insurers argued that even if Clause 8.2 meant what it said, English law does not let parties contract out of the consequences of fraud on the basis that a party should not profit from its own wrongdoing.

The Court agreed with the principle but not its application. The authorities on which the Insurers relied all deal with fraud which had been proved. The fact that the allegation was still live meant that this present case was clearly distinguishable. Further, the Court held that Clause 8.2 still could work in insurers’ favour in that if fraud is later established in accordance with the policy, the policy is voided from inception and everything paid out, including defence costs, must be returned. The clause simply determines who carries the risk in the gap between allegation and proof.

That logic extends beyond D&O insurance to any contract requiring continued performance while fraud allegations remain unresolved. The commercial difficulty is that clawing back advanced defence costs from an individual insured, even with a clear right of recovery, is often impractical, very expensive and ultimately unsuccessful.

Defence costs arising from unproven criminal allegations, including bribery, are insurable

Insurers separately argued that the defence costs were simply uninsurable because they arose from criminal bribery allegations. Again, the Court disagreed. Applying Coulson v News Group Newspapers Ltd [2012] EWCA Civ 1547, it held there is nothing objectionable about one person funding another’s defence to a criminal charge.

Bribery is serious, but seriousness does not make a defence uninsurable. If it did, as the Court put it, any D&O policy covering criminal allegations would be “somehow ineffective or impermissible“, which would be “a most surprising conclusion“. The policy itself reinforced the point as criminal fines and penalties were excluded from “Loss”, but Bribery Act 2010 exposure was expressly included in the definition of “Wrongful Act”. The Court was clear that insurability must be assessed on the basis of the allegations as they stand, not a hypothetical future conviction. An insurer that could pull funding on the strength of an allegation alone would effectively decide the outcome of the case before it was tried.

The decision of the Court was not particularly surprising. The purpose of a D&O policy is to provide cover for defence costs to individuals and sometimes those individuals face allegations of dishonesty which might, at some point, entitle insurers to decline an indemnity. The context of affording coverage to individuals is important because it is that which distinguishes a D&O policy from other liability policies and makes the final adjudication language so common.

Practical takeaways

1. Review non-avoidance and conduct exclusion wording together. An “is established by adjudication or admission” formulation may defer the insurer’s right to act until fraud is independently established or admitted, not merely restate the ordinary burden of proof.
2. Price and reserve the interim funding gap as a real cost. Even though a right of recovery survives if fraud is later established, recouping advanced defence costs from an individual insured person is often difficult in practice and should not be assumed to be likely.
3. There is no public policy bar to insuring the costs of defending unproven criminal allegations.

Conclusion

D&O insurance exists to fund an effective defence during the period of uncertainty between allegation of fraud and proof. This decision holds insurers to that purpose. It is also a reminder to revisit non-avoidance and conduct exclusion wording, and to ensure that underwriting and pricing reflect the real risk of advancing costs which may never be recovered.

To learn more about how we can assist your business in considering these risks, please contact Nathan Penny-Larter and Sam Zaozirny.