Explore our new Cybersecurity Terms Glossary, explained by our experts.
Cybersecurity Terms Glossary
Exclusions are the clauses that remove cover from a cyber policy. War, infrastructure failure, unpatched systems, and misstated controls account for most of the gap between expectation and payout.

What Are Cyber Insurance Exclusions?
Every cyber policy is defined as much by what it removes as by what it grants. Exclusions sit in their own section, often several pages long, and they are where claims are won and lost. A buyer who compares policies on limit and premium alone is comparing the wrong things.
The exclusions below appear in most wordings sold in the UK market. The specific language varies by insurer, and that variation is exactly what a broker should be asked to explain line by line.
This is the exclusion that changed the market. After the NotPetya attack of June 2017 caused an estimated $10 billion of global damage, insurers denied large claims from Merck and Mondelez on the basis that the attack was an act of war by Russia. Merck’s case ended in a settlement in January 2024 after a New Jersey appeals court ruled the traditional war exclusion did not clearly apply to a cyber event, and Mondelez settled separately.
Lloyd’s responded by requiring all standalone cyber policies written at the market to carry a state-backed cyber attack exclusion from 31 March 2023. Four model clauses, LMA5564 to LMA5567, offer different treatments, chiefly over whether cover continues for attacks on a state’s own critical infrastructure and how attribution is determined.
The practical questions for a buyer: who decides whether an attack was state-backed, what evidence is required, and whether a government statement is binding on the insurer. Attribution in cybersecurity is slow and contested, and a clause that lets the insurer decide unilaterally is worth negotiating.
Most policies exclude losses caused by failure of infrastructure outside your control: power, telecommunications, satellite, and internet backbone. Cloud outage treatment varies considerably. Some wordings exclude it, some cover it as contingent business interruption with a separate sublimit, and some cover only named providers. Given how concentrated cloud dependency has become, this clause deserves more attention than its length suggests.
A cluster of exclusions turns on your own behaviour rather than the attacker’s.
Policies pay to restore systems to their pre-incident state, not to improve them. If your recovery involves migrating to a newer platform, adding segmentation, or deploying tooling you did not previously own, the insurer will treat the difference as betterment and decline it. This is the most frequent source of surprise after a ransomware claim, because the rebuild an organisation wants is rarely the rebuild it had.
Two clauses control when an incident counts. The retroactive date excludes acts occurring before a stated date, which matters because intrusions are often discovered months after they began; the median dwell time reported by incident responders has fallen but remains long enough to cross a policy boundary. Prior knowledge excludes circumstances a responsible officer knew about before inception, including a known unremediated finding from a penetration test you failed to disclose.
Bodily injury and physical property damage normally sit outside cyber cover and belong to other lines, which creates genuine gaps for manufacturers and operators of operational technology. Intellectual property infringement, contractual penalties and liquidated damages, loss of future revenue or goodwill, and in many wordings social engineering fraud, are also excluded or moved to a small sublimit; funds transfer fraud is frequently a separate insuring clause with its own limit.
Three practical steps. Map every exclusion against your actual environment and record where cover stops, then take that list to the risk committee as accepted risk. Ask the insurer to confirm in writing how the state-backed exclusion would apply to a named scenario, such as a destructive attack on a shared service provider. And align your control evidence with the warranties, so the patching records, backup test logs, and MFA coverage reports exist before a claim rather than after.
None of this replaces the controls themselves. Our cyber risk analysis services help organisations evidence the controls insurers now treat as conditions, and understand what the policy will not carry.
It removes cover for losses arising from war and, under clauses required at Lloyd’s since 31 March 2023, from state-backed cyber attacks. The dispute over NotPetya showed how contested this can be. Check who determines attribution, what evidence is needed, whether a government statement binds the insurer, and whether attacks on critical infrastructure remain covered.
Yes, where the wording contains an unpatched software exclusion or a patching warranty. These typically exclude losses from a vulnerability that has had a vendor fix available longer than a stated period, commonly 45 or 90 days. Keep dated evidence of your remediation timelines, because the burden of showing compliance will fall on you.
Betterment is any improvement over the pre-incident state. Insurance restores what you had, so if recovery involves upgrading platforms, adding segmentation, or deploying tooling you did not own before, the insurer funds the equivalent of the old environment and declines the uplift. Budget separately for the improvements you will want to make during a rebuild.
Often only partly. Many policies treat social engineering and funds transfer fraud as a separate insuring clause with a much lower sublimit than the main policy aggregate, and some exclude it entirely, leaving it to a crime policy. If payment fraud is a realistic exposure for your finance function, check the sublimit and the conditions attached to it specifically.
You May Be Interested In These
We use a few cookies that are needed to make this website work, and they are always on. We would also like to set analytics cookies, to see which pages are useful; functional cookies, so that our live chat works; and advertising cookies, to measure our campaigns. We will not set any of those until you say yes, and you can change your mind at any time from the Cookie Settings link in the footer.
Choose the cookies you are happy for us to use. Strictly necessary cookies are always on; everything else is off until you turn it on.
These cookies keep the website working: they hold your security checks, your form submissions and the choice you make here. The site cannot run without them, so they are always on.
These let our live chat open and remember your conversation, and let the newsletter sign-up work. Switch them off and the chat window will not appear.
These tell us which pages are read and how people find us, so that we can improve the site. The reports are about groups of visitors, never about you personally.
These let us measure our advertising and show you our adverts on other websites, such as LinkedIn, Google and Meta. Switch them off and you will still see adverts, but they will be less relevant.