A regional firm of consulting engineers receives a complaint letter from a developer client about alleged design errors on a residential scheme. The principal reads it, assumes it will blow over, and files it. Eleven weeks later the developer issues a Letter of Claim quantifying losses at £2.6 million. The engineers’ PI policy contains a clause requiring notification of any “circumstance which may give rise to a claim” as soon as reasonably practicable, expressly stated to be a condition precedent to the insurer’s liability. Cover for the entire matter is now in serious doubt — not because the firm caused the loss negligently, but because of how the policy term is classified.
This is the territory where warranties, conditions precedent and “mere” conditions live. The labels look interchangeable on the page. The consequences of breach are entirely different. After the Insurance Act 2015 the distinction has, if anything, become more important, not less — because the Act changed what happens when a warranty is breached, but left the law on conditions precedent essentially untouched.
This article explains the four categories brokers and insureds need to recognise, how the courts construe them, the case law that matters, and what to do at placement and at notification to avoid losing cover on a technicality.
Why the distinction matters
The label a policy uses for a term governs the remedy the insurer has when that term is broken. The four categories, and the consequences of breach, are:
- Warranty — under s.10 of the Insurance Act 2015, breach suspends the insurer’s liability from the moment of breach until the breach is remedied. Cover resumes thereafter. Losses occurring during the suspension period are not covered.
- Condition precedent to the insurer’s liability under the policy — breach means the insurer has no liability for the affected claim. The policy continues, and other unaffected claims remain payable.
- Condition precedent to the contract — the policy never comes into force at all. There is no insurance. Premium is generally returnable.
- Mere condition — breach gives rise to a claim in damages only. The insurer must still pay the claim but can set off any loss caused by the breach. There is no right to avoid, terminate, suspend or refuse cover.
Four labels, four different outcomes. The same notification clause can fall into any of the first, second or fourth category depending on how it is drafted. That is why the classification exercise has to happen at placement, not after a loss.
Warranties post-Act: s.10 and s.11
Before the Insurance Act 2015, a breach of warranty automatically discharged the insurer from all liability from the date of breach — even if the breach was remedied, even if it was trivial, and even if it had no connection to the loss. The classic illustration was The Good Luck [1992] 1 AC 233.
The Act made two fundamental changes.
Section 10 abolished the rule of automatic discharge. A breach of warranty now suspends the insurer’s liability from the time of breach until the breach is remedied. Once remedied, cover resumes for losses occurring thereafter. Losses occurring during the suspension period remain outside cover.
Section 11 provides that where a term (including a warranty) is intended to reduce the risk of loss of a particular kind, at a particular location, or at a particular time, the insurer cannot rely on non-compliance to exclude or limit its liability if the insured shows that the non-compliance could not have increased the risk of the loss that actually occurred in the circumstances in which it occurred.
The interplay matters: a warranty post-Act has suspensive rather than terminal effect, and the suspension may itself be unavailable if s.11 applies. This is dealt with in detail in our companion article on warranties.
For the purposes of this article, the takeaway is narrower: a warranty in a modern policy is no longer the nuclear option it used to be. Conditions precedent are now, in many policies, the more dangerous term — because s.10 does not soften their effect.
Condition precedent to the insurer’s liability
This is the workhorse category for claims-handling disputes. Most disputes about notification, cooperation, claim form requirements, document preservation and prompt mitigation turn on whether the breached term is a condition precedent to the insurer’s liability.
The key features the courts have repeatedly applied are:
- Strict construction. Conditions precedent are construed against the insurer (contra proferentem) and the words must be clear.
- Express identification or unambiguous effect. The clause must either be expressly stated to be a condition precedent or must, on a fair reading of the wording, unambiguously have that effect.
- Causation is irrelevant to the breach itself. Unlike a mere condition where the insurer must show loss flowing from the breach to obtain damages, a condition precedent to liability “bites” on breach: the affected claim falls out of cover.
In Aspen Insurance UK Ltd v Pectel Ltd [2008] EWHC 2804 (Comm) the court considered a notification clause expressly drafted as a condition precedent and confirmed that, where the words are clear, the consequences are uncompromising: failure to notify within the contractual period defeats cover for that claim, irrespective of whether the insurer has suffered prejudice.
In Friends Provident Life & Pensions Ltd v Sirius International Insurance Corporation [2005] EWCA Civ 601 the Court of Appeal addressed what a clause must say to be classified as a condition precedent. The court refused to imply condition-precedent status where the wording did not clearly demand it. A general “Conditions” heading was not enough. Clear words — typically “it is a condition precedent to the liability of the insurer that …” or an express labelling of the term as a condition precedent — are required.
The Court of Appeal in Aspen Insurance UK Ltd v Adana Construction Ltd [2015] EWCA Civ 176 reinforced the strictness of construction in a construction-PI context, examining how a notification provision interacts with the insurer’s right to refuse cover and emphasising that ambiguity is resolved against the insurer.
Zurich Insurance plc v Maccaferri Ltd [2016] EWCA Civ 1302 is essential reading on notification triggers. The clause required notification “as soon as possible after the occurrence of any event likely to give rise to a claim”. The Court of Appeal held that “likely to give rise to” required an objective assessment based on what the insured knew at the time; it was not enough that, viewed with hindsight, an event turned out to give rise to a claim. The decision is helpful to insureds because it limits the universe of notifiable circumstances to those that, on the information available, are objectively likely to result in a claim. It also illustrates that the careful construction of trigger language can be decisive.
The lesson for brokers is that any clause your client must comply with on notification — and any clause requiring cooperation, document production, or particular conduct during a claim — needs to be checked against the heading “is this a condition precedent?”. If it is, the trigger must be understood and a compliance process built around it.
Condition precedent to the contract
This category is much rarer in commercial classes but does occur. A condition precedent to the contract is a term the satisfaction of which is required for the policy to come into existence at all. If it is not met, there is no contract — no cover, no claim, generally a return of premium.
Examples include premium-warranty arrangements drafted as conditions precedent to inception, or specific risk-management requirements that must be in place before cover attaches (for instance, a sprinkler system in commission as a condition of inception rather than a condition of continuing cover).
The drafting tell is language pointing to inception or to the formation of the contract — for example, “this contract shall not come into force unless…” or “it is a condition precedent to the formation of this contract that…”.
Mere conditions
A mere condition is a term that is neither a warranty nor a condition precedent. Breach is actionable in damages by the insurer, but does not allow the insurer to avoid the contract, terminate cover, suspend liability or refuse to pay the affected claim. The insurer must prove loss caused by the breach to recover damages, and any recovery is by way of set-off against the claim payment.
Many policy terms are properly classified as mere conditions even though insurers sometimes argue otherwise. Examples include general housekeeping requirements that are not expressly elevated to condition-precedent status, certain reporting obligations not tied to the affected claim, and assistance obligations that are not in terms a precondition to payment.
The practical point: if the contract does not say “condition precedent”, and does not use words that unambiguously create that effect, the courts will not lightly imply it. Friends Provident is authority for this proposition and brokers can use it to push back when an insurer treats a non-precedent term as if it were one.
Does s.11 apply to conditions precedent?
Yes. This is important and frequently misunderstood.
Section 11 is not limited to warranties. It applies to any “term” of the contract — including conditions precedent — that defines the risk by reference to the type of loss, the location of the loss, or the time of the loss. The Law Commissions’ Joint Report (2014) confirmed Parliament’s intention that the protection should not be circumvented by relabelling a warranty as a condition precedent.
Where a condition precedent is properly characterised as a risk-mitigation term within s.11 — for example, a requirement to keep premises locked at night, which is about a loss “at a particular time” — the insurer cannot rely on the breach to defeat the claim if the insured shows the breach could not have increased the risk of the actual loss in the circumstances in which it occurred.
Section 11 does not assist with notification, cooperation or claims-handling conditions precedent. Those clauses are not risk-mitigation terms; they are claims-procedure terms. A late-notification breach is not made good by saying it could not have increased the risk of the underlying loss. The drafting target of s.11 was clauses that try to use procedural levers to escape risks the underwriter has actually priced for.
Worked example: a notification breach
Take the engineering firm in the opening scenario, and assume the policy wording reads: “It is a condition precedent to the liability of the Insurer that the Insured shall give written notice to the Insurer as soon as reasonably practicable of any circumstance of which the Insured becomes aware which may give rise to a claim.”
If the term is, as drafted, a condition precedent to the insurer’s liability, the eleven-week delay between receiving the complaint letter and notifying — assuming the complaint letter triggered the obligation — defeats cover for the matter. Insurer prejudice is irrelevant. Section 11 does not assist because the clause is a claims-procedure term.
Had the same clause been drafted as a warranty (“the Insured warrants that notification shall be given within 14 days”), s.10 would apply: cover would be suspended during the period of breach. But where the breach is by definition a one-off historical failure, “remedying” the breach is conceptually difficult, and the practical outcome would often still be loss of cover for the affected matter.
Had the clause been a mere condition, the insurer would have to pay the claim but could set off any loss caused by the delay — typically the additional defence costs caused by late involvement, or any settlement uplift attributable to a worse litigation position.
Had it been a condition precedent to the contract, the policy would never have come into force at all — which is not realistic on a notification clause, but illustrates the spectrum.
Same facts. Same delay. Four different outcomes.
How brokers should review policy wordings
A condition-precedent review at placement should focus on:
- The “Conditions” or “Conditions Precedent” section of the policy schedule and wording, where insurers usually consolidate these terms.
- Carve-outs within the General Conditions that say a sub-set of conditions are precedents.
- Defined terms — sometimes “Condition Precedent” is a defined expression linked back to a numbered list.
- Notification clauses — read every word: trigger (claim, circumstance, occurrence, event), timing language (immediately, as soon as reasonably practicable, within X days), and consequence (expressly stated as a condition precedent or not).
- Co-operation, assistance and document-preservation obligations.
- Claims control and conduct-of-defence clauses, which in liability policies are often conditions precedent.
- Premium-warranty clauses and any inception-conditional language.
The output of the review should be a one-page client brief identifying each condition precedent, the trigger, the time limit, and the action required. That brief is the document that prevents the Aspen v Pectel outcome from happening to your client.
What to do at notification
Notification is the most common point at which a condition-precedent breach occurs. The discipline at notification is straightforward but easily lost when an insured is busy or hopeful that a problem will resolve itself:
- Notify on awareness, not certainty. Most modern notification clauses trigger on a circumstance “which may give rise to” or is “likely to give rise to” a claim. Maccaferri sets the threshold for “likely”. For “may give rise to”, the threshold is lower still.
- Notify in writing to the correct address. Verbal notification to the broker is not always sufficient; many clauses require written notice to a named address.
- Notify even if uncertain. A precautionary notification preserves cover. A non-notification on the view that the claim will go away cannot be cured later if the policy then expires before the claim is made and the next policy excludes prior known matters.
- Document the notification. Keep proof of dispatch and acknowledgement of receipt.
- Escalate internally. Build a notification protocol into the insured’s incident-response and complaints procedures so that the people who first see a problem know to refer it.
The eleven-week delay in our opening scenario is not unusual. Most late notifications begin with someone telling themselves a problem will sort itself out. The cost of that judgement call is whatever the eventual claim turns out to be.
Frequently Asked Questions
Is every clause in the “Conditions” section of a policy a condition precedent? No. A clause is a condition precedent only if it is expressly identified as one or unambiguously has that effect. Friends Provident v Sirius [2005] EWCA Civ 601 confirms that a general “Conditions” heading is not enough.
Has the Insurance Act 2015 changed the law on conditions precedent? The Act did not codify or reform the law on conditions precedent directly. It changed the consequences of breach of warranty (s.10) and introduced s.11 on terms not relevant to the actual loss. Section 11 can apply to conditions precedent that are risk-mitigation terms, but not to claims-procedure terms like notification clauses.
Does an insurer need to show prejudice to rely on breach of a condition precedent? No. That is the central difference between a condition precedent to liability and a mere condition. Aspen v Pectel confirms that prejudice is not required.
What is the difference between a warranty and a condition precedent after the Act? A warranty breach now suspends cover under s.10 until remedied. A breach of a condition precedent to liability defeats cover for the affected claim. A warranty breach can sometimes be cured; a one-off breach of a notification condition precedent generally cannot.
Does s.11 protect against notification breaches? No. Notification is a claims-procedure term, not a risk-mitigation term, and falls outside s.11.
What is the practical difference between “as soon as possible” and “as soon as reasonably practicable”? “As soon as reasonably practicable” gives the insured more latitude and is judged with reference to what was reasonable in the circumstances. “As soon as possible” is stricter. Brokers should push for the former.
Can a condition precedent be contracted out of in favour of the insured? A clause more favourable to the insured than the default position is permissible — there is no obstacle to softening a condition precedent into a mere condition by agreement, and brokers should consider it where appropriate.
Should brokers be flagging conditions precedent in the demands and needs statement? Yes. FCA conduct rules require that significant terms — and condition precedents are by their nature significant — are drawn to the client’s attention. Documenting that briefing is best practice.
Related articles in this series
- The Insurance Act 2015: an overview from the PI broker’s perspective
- The duty of fair presentation under the Insurance Act 2015
- The material circumstance test
- Warranties under the Insurance Act 2015 reform
- Terms not relevant to the actual loss (s.11)
- Remedies for breach under the Insurance Act 2015
- Contracting out of the Insurance Act 2015
- Late payment of claims and s.13A
- The Insurance Act 2015 and PI claims
Reviewed by the Apex Insurance Brokers technical team, May 2026. Apex Insurance Brokers Ltd is authorised and regulated by the Financial Conduct Authority (firm reference 724952) and is registered in England and Wales (Companies House 07014570). This article is general guidance and is not legal advice; insureds should take advice on specific facts.