Warranties Reform under the Insurance Act 2015

A Bristol-based logistics operator we placed cover for some years ago discovered, at three in the morning on a Sunday, that the intruder alarm at one of its warehouses had gone offline. The maintenance contractor reset the system on Monday. Three weeks later, a small electrical fire broke out in the same building. Before the Insurance Act 2015, that operator would have been arguing — possibly in court — about whether an unrelated alarm breach weeks earlier had quietly terminated their property cover. After the Act, the answer is far cleaner: the alarm warranty was suspended while the alarm was down, was reinstated when it came back online, and the fire claim is payable.

That shift — from automatic, permanent discharge of cover to a temporary, suspensive remedy — is what section 10 of the Insurance Act 2015 delivers. For commercial insureds and the brokers who place their cover, it is one of the most significant changes in modern insurance law. This article explains the old rule, the new rule, the exceptions, and what brokers should be doing at placement and at the point a breach is identified.

The pre-2015 position: automatic discharge under the Marine Insurance Act 1906

For more than a century the law on warranties was governed by section 33(3) of the Marine Insurance Act 1906, which applied not only to marine cover but, by judicial extension, to non-marine insurance as well. The provision was uncompromising: a warranty “must be exactly complied with, whether it be material to the risk or not”, and breach discharged the insurer from liability “as from the date of the breach of warranty”.

The leading modern authority on the harshness of that rule was the House of Lords decision in Bank of Nova Scotia v Hellenic Mutual War Risks Association (Bermuda) Ltd (The Good Luck) [1992] 1 AC 233. Lord Goff confirmed that breach of a promissory warranty produced automatic termination of the insurer’s liability from the date of the breach — there was no need for the insurer to elect to terminate, and the breach could not be cured. The contract was not avoided ab initio; rather, the insurer’s obligations simply fell away from the moment of breach onwards.

The consequences were stark. A trivial, immaterial, or even rectified breach of warranty could leave the insured with no cover at all for a loss that occurred long after the breach had been put right, and which had nothing whatever to do with the breach. The operator whose alarm was offline for one weekend could lose flood cover for the following winter. The factory whose sprinkler test was a week late could lose theft cover for a year. Few outcomes in English commercial law were more frequently criticised — by judges, by the Law Commissions and, eventually, by Parliament.

The new regime: s.10 of the Insurance Act 2015

Section 10 of the Insurance Act 2015 rewrites the law of warranties for all contracts of insurance entered into or varied on or after 12 August 2016. The key reforms operate at two levels.

Section 9 — basis of contract clauses abolished

Before turning to s.10 it is worth noting s.9, which sits alongside it. Basis of contract clauses — wording that converted every statement made in a proposal form into a warranty regardless of its materiality — had already been outlawed in consumer insurance by the Consumer Insurance (Disclosure and Representations) Act 2012 (CIDRA). Section 9 of the 2015 Act extends that prohibition to non-consumer insurance. Any term in a non-consumer policy or in any other contract with the insured (including the proposal form) which has the effect of converting a representation into a warranty is, to that extent, of no effect.

Brokers should still occasionally see legacy declarations and proposal-form “declarations of truth” purporting to incorporate every answer as a warranty. They are void. If wording of that kind survives in a policy schedule or in a slip endorsement, it should be challenged and removed at renewal.

Section 10(1) — abolition of automatic discharge

Section 10(1) abolishes any rule of law that breach of a warranty (express or implied) in a contract of insurance results in the discharge of the insurer’s liability under the contract. The automatic-discharge consequence of The Good Luck is gone.

Section 10(2) — the suspensive rule

In its place, s.10(2) introduces a suspensive effect. An insurer has no liability under the contract in respect of any loss occurring, or attributable to something happening, after a warranty in the contract has been breached but before the breach has been remedied. The corollary, in s.10(4)(b), is that the insurer’s liability resumes once the breach has been remedied.

In practice, this means warranties now operate like a switch. While the insured is in breach, cover is off. The moment the breach is put right, cover is back on. A loss occurring during the period of breach is not covered; a loss occurring after the breach has been remedied is covered, even if there was an earlier, unrelated breach.

Sections 10(3)–(6) — when remediation does not restore cover

The Act is realistic about the fact that not every breach can be remedied. Section 10(5) sets out two cases in which a breach is treated as remedied for the purposes of s.10(2):

• where the warranty required something to be done (or not done), or a condition to be fulfilled, or something to be the case, by a particular time, and the risk to which the warranty relates later becomes essentially the same as before the breach; or
• in any other case, where the insured ceases to be in breach.

Section 10(6) deals expressly with time-specific warranties: where a warranty required compliance by a particular time, and that time has passed, the breach cannot be remedied for the purposes of the Act. A warranty that the insured “will install a sprinkler system by 1 October” is, after 1 October, simply broken — there is no cure for the period of breach (though installation will of course stop the suspension running forwards).

Section 10(3) preserves the insurer’s position for losses occurring during the period of suspension: the insurer is not liable for those, and that is so even if the breach is later remedied.

The combined effect is straightforward in the simple case and nuanced in the harder ones:

• Continuing warranties (e.g. “the insured warrants that the intruder alarm will be maintained in full working order at all times”) suspend cover for the period of breach and resume cover when the alarm is fixed.
• Warranties as to past or existing facts (e.g. “the insured warrants that as at inception the premises are fitted with a working sprinkler system”) that turn out to be untrue at inception are, in practice, not capable of remedy — though brokers should consider whether the wording is really a warranty at all, or a representation that should be analysed under the duty of fair presentation.
• Time-specific warranties are capable of being put right going forward but the period of breach cannot be retroactively cured.

A worked example: the warehouse alarm

Take the Bristol operator described at the top of this article. The policy includes a warranty that “the insured shall maintain the intruder alarm system in full working order at all times and shall report any failure to the insurer within 24 hours”.

Scenario A — the alarm fails on Saturday, is repaired on Monday morning, and a fire occurs three weeks later. Under the old law, the insurer could (and frequently would) argue automatic discharge from the date of breach. Under s.10 the breach was remedied on Monday. Cover resumed. The fire claim is payable.

Scenario B — the alarm fails on Saturday and a burglary occurs on Sunday night, while it is still inoperative. Under both the old and the new law, that loss falls within the period of suspension and is not covered. Section 10(2) and (3) preserve the insurer’s position for losses during the suspension window.

Scenario C — the alarm fails on Saturday, is not repaired, and there is then a flood claim entirely unrelated to the alarm. Here we move from s.10 into s.11 (terms not relevant to the actual loss), addressed in the next article in this series. Even though the insured is technically in breach of the alarm warranty when the flood occurs, the alarm warranty is directed at the risk of theft, not flood — and s.11 may prevent the insurer from relying on the breach to refuse the flood claim.

How insurers and policy drafters have responded

Since 2016 we have seen a number of drafting responses from the London market and from regional commercial insurers.

• Recharacterisation as conditions. Some insurers have stopped using “warranties” altogether for matters that previously sat in warranty wording, in favour of conditions or conditions precedent. The choice of label matters: a true warranty engages s.10; a condition precedent to liability engages different remedies; a mere condition gives rise only to a damages claim. The change of label can shift outcomes, particularly where s.11 is in play.
• Tightly defined remediation steps. Where warranties remain, sensible insurers now include express, practical remediation steps so that the insured (and the broker on notification) can identify exactly when the breach is considered remedied for s.10 purposes.
• Risk-defining language. Some insurers have attempted to characterise warranties as terms “defining the risk as a whole” in an effort to take them outside s.11. The substance, not the label, matters: see the discussion in the next article in this series.
• Contracting out. In non-consumer contracts only, parties may contract out of parts of the Act under Part 5 (transparency requirements apply). Brokers should be alert to contracting-out wording on schedules and endorsements.

The construction of warranties: HIH v New Hampshire

The pre-Act case law on the construction of warranty wording remains relevant. The Court of Appeal in HIH Casualty and General Insurance Ltd v New Hampshire Insurance Co [2001] EWCA Civ 735 confirmed the orthodox approach: the court will look at the nature, purpose and language of the term and ask whether the parties intended the strict consequences of warranty status. Not every term labelled “warranty” is a warranty in the technical sense — and conversely, a term not so labelled may still be a warranty if its substance fits.

Post-Act, the consequences are different but the construction exercise is the same. A broker reviewing a policy should ask, of any term that might be a warranty:

• Is it a promise about a present or future state of affairs (warranty), or a precondition to liability (condition precedent), or a mere term sounding in damages?
• If it is a warranty, is it continuing or as to past or existing facts?
• Is it time-specific (in which case s.10(6) bites)?
• Could compliance with it tend to reduce the risk of loss of a particular kind, at a particular location, or at a particular time? If yes, s.11 is also in play.

Interaction with s.11

Section 10 deals with the mechanism of warranty breach — what happens to cover when a warranty is broken. Section 11 deals with the relevance of the breach to the loss — even if cover is technically suspended, can the insurer rely on the breach to refuse a loss that the breach could not have caused?

The two sections work together. A broker advising on a warranty breach should always ask both questions. The next article in this series (Terms Not Relevant to the Actual Loss) addresses s.11 in detail.

What brokers should look for at placement

A practical broker checklist for warranty review at placement and renewal:

1. Identify every warranty in the wording. Read the schedule, the policy wording, every endorsement and the proposal form. Warranties are sometimes labelled “warranty”, sometimes labelled “condition precedent”, and sometimes buried in a declaration.
2. Strike out basis-of-contract clauses. Anything in the proposal form, declaration of truth or policy that converts representations into warranties in a non-consumer contract is void under s.9. Mark it up and ask the insurer to remove it from the wording rather than leave it lying on the file.
3. Ask whether each warranty is necessary. Some warranties exist out of historic drafting habit rather than current risk concern. Negotiate to convert non-critical warranties into conditions or risk-defining terms.
4. Test for practicality of remediation. Can the insured actually comply? If the warranty is a continuing one (e.g. “the alarm will be tested weekly”), is there a documented process and a named owner? If not, the warranty is likely to be breached at some point — and although s.10 cushions the consequences, suspension during the breach period is still a real exposure.
5. Identify time-specific warranties. Any “by [date]” warranty is a hard deadline under s.10(6). Diarise it with the client.
6. Check for s.11 protection. For each warranty, ask whether it is risk-specific (kind, location or time of loss) or risk-defining as a whole. The former gives the insured s.11 protection; the latter does not.
7. Watch contracting-out wording. Under Part 5 the parties can contract out of s.10 in non-consumer contracts, subject to transparency requirements. Any such clause should be flagged to the client in writing and considered against the price and the cover.

What brokers should do when a breach is notified

When a client notifies a breach (or a near-breach), the broker should:

• Establish the date and time the breach began, and the date and time it was remedied (or whether it has been remedied at all).
• Establish whether any loss occurred during the suspension window.
• For losses outside the suspension window, document the s.10(2) and (4)(b) position and put the insurer on notice that the suspension has ended.
• For losses inside the suspension window, consider s.11 — was the breach causally relevant to the loss?
• Where the warranty is time-specific, identify whether s.10(6) applies and assess whether the breach is now incapable of remedy for the purposes of the Act.

FAQs

1. Does the Insurance Act 2015 apply to all my commercial policies? It applies to non-consumer contracts of insurance entered into, varied or renewed on or after 12 August 2016, except where the parties have lawfully contracted out under Part 5. Most UK commercial covers are now within scope.

2. What is the difference between a warranty and a condition precedent? A warranty, post-Act, suspends cover under s.10(2) for the period of breach. A condition precedent to liability prevents the insured from recovering for a particular loss if the precedent is not satisfied. Both are stricter than a mere policy condition (which sounds only in damages). The label is indicative but not decisive — substance matters.

3. Can a warranty breach be remedied? Yes, in most cases. Under s.10(2) and s.10(5), cover resumes once the breach is remedied. The exception is time-specific warranties under s.10(6), where the breach cannot be remedied for the period before the time expired.

4. Are basis of contract clauses still enforceable? No. Section 9 abolishes them in non-consumer contracts; CIDRA already abolished them in consumer contracts. Any such clause is of no effect and should be removed from policy wording at renewal.

5. If I’m in breach of a warranty and then have a loss unrelated to the warranty, am I covered? Possibly. Section 10 suspends cover during the breach period, but section 11 may prevent the insurer from refusing a loss that the breach could not have caused. See our s.11 article.

6. Can my insurer contract out of s.10? In non-consumer insurance only, and only if the transparency requirements in Part 5 are met. Contracting out should be a red flag for the broker and the insured, and should be explained clearly in writing.

7. Does s.10 apply to reinsurance? Yes — reinsurance contracts are within scope of the Act, although contracting-out is more common in that market.

8. What records should I keep to evidence remediation of a breach? Dated maintenance reports, contractor invoices, photographs, written confirmation from the contractor of the date and time of remediation, and a contemporaneous note to the broker. These will be the documents an insurer (or, ultimately, a court) will look at.

Related articles in this series

1. The Insurance Act 2015 — a broker’s overview
2. The duty of fair presentation
3. The material circumstance test
4. Warranties reform under the Insurance Act 2015 (this article)
5. Terms not relevant to the actual loss — s.11
6. Remedies for breach of the duty of fair presentation
7. Contracting out under Part 5
8. Condition precedent vs warranty post-Insurance Act
9. Late payment of claims under the Enterprise Act 2016
10. The impact of the Insurance Act 2015 on 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.