Limitation of liability clause
How liability caps work, what carve-outs really mean, and where the hidden exposure is.
What is a limitation of liability clause?
A limitation of liability clause caps the maximum amount one party can claim from the other. Most combine two things: a liability cap (a ceiling on total exposure) and exclusions (categories of loss that are off the table entirely, like consequential damages).
This is the clause that tells you the financial worst case. A supplier on a £50,000/year contract needs to know whether a catastrophic failure means a £50,000 claim or an unlimited one.
Relevant legislation
Most jurisdictions allow parties to limit liability by contract, but with constraints:
- UK: Unfair Contract Terms Act 1977 (UCTA). You can't exclude liability for death or personal injury from negligence. Other exclusions must pass a "reasonableness" test. If a court finds the clause unreasonable, it's struck down entirely. No cap at all.
- US: varies by state. Courts apply unconscionability doctrine. Some states have specific statutes limiting what can be excluded (e.g. California Civil Code § 1668 voids certain liability waivers for negligence).
- Australia: Australian Consumer Law. Consumer guarantees can't be excluded. Limitations in B2B contracts must not be "unfair" in standard form contracts.
The specifics differ, but the pattern is the same: courts won't enforce limitations they consider unreasonable or unconscionable, and if they strike one down, the party usually ends up with no cap at all.
What to look for
The cap amount is the starting point. Is it 1x annual fees? 2x? Something lower? A cap of "fees paid in the preceding month" looks like a limitation but effectively eliminates any meaningful remedy. The standard range for B2B contracts is 1x to 2x the annual contract value.
The cap basis matters just as much. Does it apply per incident, per year, or for the entire contract term? A per-incident cap with no aggregate means multiple incidents can each generate claims up to the full cap. A single aggregate cap for the entire term means one large claim can exhaust everything.
Check the exclusions. Most contracts exclude consequential or indirect damages (lost profits, business interruption, reputational harm). Whether that's acceptable depends on context. For commodity services, it's standard. For a hosting agreement where the main risk is downtime, excluding consequential damages removes the customer's remedy for the exact loss they're most likely to suffer.
Check the carve-outs, meaning obligations excluded from the cap. Common ones: indemnification, IP infringement, breach of confidentiality, data protection (GDPR fines). Each carve-out creates uncapped exposure. A contract with a $100,000 cap but unlimited carve-outs for indemnification and data protection is, in practice, an uncapped contract.
Common pitfalls
Caps so low they provide no real remedy. Think $1, or "fees paid in the preceding month." These create a false sense of protection. Courts may strike them down, but that takes litigation, which is expensive and uncertain.
Too many carve-outs can make the cap meaningless. If every major obligation sits outside the cap, the cap only covers a narrow residual category. The stated cap and the actual exposure are completely different numbers.
A contract with no limitation clause at all exposes both sides to theoretically unlimited liability. It's rarer than you'd think, but it happens, especially in bespoke agreements drafted in a hurry.
Overly aggressive limitations can backfire. In many jurisdictions, if a court finds a limitation unreasonable or unconscionable, it doesn't substitute a "fair" cap. The clause is gone entirely. The party who pushed for the lowest possible cap ends up with no cap at all.
Example clause
"The total aggregate liability of either party under this Agreement shall not exceed 150% of the fees paid or payable by the Customer in the 12 months preceding the claim. Neither party shall be liable for any indirect, consequential, or special losses, or for loss of profit, revenue, data, or business. The limitations in this clause do not apply to liability under clause 5 (Confidentiality) or clause 10 (Indemnification)."
Frequently asked questions
Can you completely exclude liability in a contract?
Not in most jurisdictions. Courts apply some form of reasonableness, unconscionability, or fairness test. A clause that tries to exclude everything will almost certainly be struck down, and then you have no limitation at all. The practical approach is to limit liability to a reasonable amount rather than trying to eliminate it.
What is a reasonable liability cap?
1x to 2x the annual contract value is the common range in B2B contracts. Courts look at bargaining power, availability of insurance, and whether the other party could have gone elsewhere. Higher-risk services (data processing, critical infrastructure) normally justify higher caps.
What are consequential damages?
Losses that don't flow directly from the breach but arise as a secondary consequence: lost profits, business interruption, reputational harm. Most contracts exclude them. That's usually fine, but check whether the exclusion removes the remedy for the loss your client is most likely to suffer.
How Clara helps
Clara identifies limitation of liability clauses and analyses the cap amount, cap basis, exclusions, and carve-outs. It flags asymmetric caps, unusually low limits, and missing limitation clauses so you can assess the actual financial exposure.
See how Clara flags this in your contracts
Upload a contract and get AI-powered risk analysis in about 60 seconds.