Limitation of Liability Clauses in Indian Contracts: What Actually Holds Up

A limitation of liability clause sets a ceiling on how much one party has to pay the other if something goes wrong, and it usually also carves out entire categories of loss, like lost profits, that neither party can claim at all. The one thing most people get wrong: they read the cap number and stop there. The real risk sits in the carve-outs, the exclusions, and whether the cap is even enforceable if it was never really negotiated. This guide (published by Adira, which makes contract review and CLM software, so we have a commercial stake in you understanding contracts well, but this explainer stands on its own) walks through what a limitation of liability clause does in an Indian contract, what the law says about when a cap can be struck down, and what to check before you sign one.

Plain meaning

A limitation of liability (LoL) clause does three things, usually in the same paragraph. First, it caps the total amount either party can be made to pay under the contract, often expressed as a multiple of fees paid in a set period, such as "12 months' fees" or "the total contract value." Second, it excludes whole categories of damage from being claimed at all, most commonly "indirect, incidental, consequential, special or punitive damages" and, often, lost profits, regardless of the cap. Third, it usually lists carve-outs, situations where the cap and exclusions do not apply, such as fraud, gross negligence, breach of confidentiality, IP infringement, or a party's indemnity obligations.

Without this clause, a party in breach is exposed to whatever Sections 73 and 74 of the Contract Act otherwise allow, which has no fixed upper limit beyond what a court finds was a natural, foreseeable consequence of the breach. The LoL clause replaces that open-ended exposure with a number both sides agreed in advance, a legitimate commercial purpose. The problem is when the number is so low, or the clause so one-sided, that it stops looking like risk allocation and starts looking like an escape from real consequence.

Who it protects and what triggers it

On paper, most LoL clauses are drafted as mutual, both parties get the same cap and exclusions. In practice they are rarely symmetric in effect. A software vendor with a cap of "12 months' fees" is protected against a customer's claim for losses from a platform outage, while the customer's own liability, usually just paying invoices, was never near that figure anyway. The clause reads as balanced but constrains one side's exposure far more than the other's.

The trigger is any claim for damages from breach or negligence, or, depending on drafting, any claim "arising out of or in connection with" the agreement. That phrase matters. A narrow clause caps damages for breach of specific obligations. A broad one can sweep in tort claims, statutory claims, and claims unrelated to the promise that was broken.

What to look for

Four mechanics decide whether an LoL clause is fair or a trap, and you will miss all of them if you only look at the headline cap figure.

1. Direct vs indirect/consequential damages. Indian courts do not treat "consequential" as a fixed term of art; contracts define it themselves. Excluding "indirect and consequential loss" is very different from excluding "any loss of profit, revenue, business, or goodwill, whether direct or indirect." The second can exclude losses most people would call direct, like lost revenue from a failed integration, just by labelling them.
2. The cap number, and what it is a multiple of. "12 months' fees" on a contract with a large one-time fee and small ongoing fees can mean a tiny cap relative to real exposure. Check whether the cap is fees in the 12 months before the claim, fees over the contract's life, or a flat sum. These differ a lot.
3. Carve-outs from the cap. The standard, defensible list is fraud, gross negligence, confidentiality breach, IP infringement, and indemnity obligations. A clause with no carve-outs, meaning even fraud is capped at 12 months' fees, is a serious red flag.
4. A super-cap for carved-out claims. Some contracts leave carve-outs technically "uncapped," which in practice means unlimited. A better-drafted contract sets a separate, higher super-cap, say 3x to 5x fees, so exposure is large but not infinite.

The Indian position: Section 23 and Sections 73-74

Two different parts of the Indian Contract Act, 1872 bear on a limitation of liability clause, and they do different jobs.

Sections 73 and 74 set the default rule for compensation if the contract is silent, or if you are testing whether a stated cap looks reasonable. Section 73 says:

"When a contract has been broken, the party who suffers by such breach is entitled to receive, from the party who has broken the contract, compensation for any loss or damage caused to him thereby, which naturally arose in the usual course of things from such breach, or which the parties knew, when they made the contract, to be likely to result from the breach of it. Such compensation is not to be given for any remote and indirect loss or damage sustained by reason of the breach." Source: Section 73, Indian Contract Act, 1872

Notice that Section 73 already excludes "remote and indirect" loss, before any clause does anything. An LoL clause excluding indirect and consequential damages largely restates what the statute already does. Where it adds real value for the drafting party is the cap on direct damages, and in stretching "indirect" wider than a court might on its own. Section 74 works alongside this where the contract states a sum payable on breach: a court awards reasonable compensation not exceeding that sum, not the number automatically.

Section 23 is the provision that can defeat an LoL clause altogether, by declaring it void, if it "is of such a nature that, if permitted, it would defeat the provisions of any law," or if the Court regards it "as opposed to public policy":

"The consideration or object of an agreement is lawful, unless it is forbidden by law; or is of such a nature that, if permitted, it would defeat the provisions of any law; or is fraudulent; or involves or implies, injury to the person or property of another; or the Court regards it as immoral, or opposed to public policy. In each of these cases, the consideration or object of an agreement is said to be unlawful. Every agreement of which the object or consideration is unlawful is void." Source: Section 23, Indian Contract Act, 1872

This is what lets a court strike down a term as void, not just interpret it narrowly. It rarely happens for negotiated commercial contracts between businesses of roughly equal size. It becomes a live risk when the LoL clause is imposed on a party with materially weaker bargaining power, take it or leave it, and drafted so one-sidedly that it looks less like risk allocation and more like an escape from real accountability.

A named Indian case: Central Inland Water Transport Corporation v Brojo Nath Ganguly

In Central Inland Water Transport Corporation Ltd v Brojo Nath Ganguly, 1986 AIR 1571, the Supreme Court struck down a service rule letting a government-owned corporation terminate a permanent employee at will, without reason, on notice. The rule sat in a standard-form contract the employee had no real power to negotiate. The Court held that such a clause, imposed on parties with grossly unequal bargaining power, is void under Section 23:

"Such a contract or such a clause in a contract ought, therefore, to be adjudged void under section 23 of the Indian Contract Act, as opposed to public policy."

Because the employer was a government company, falling within "the State" under Article 14 of the Constitution, the Court also held the arbitrary termination power unconstitutional, an angle specific to state and state-linked entities rather than purely private parties.

This case is about an employment rule, not a commercial LoL clause, and Indian courts have not used it to routinely strike down negotiated B2B liability clauses between parties of comparable size. What it establishes, and why it matters here, is the underlying doctrine: Section 23 absorbs a doctrine of unconscionability, and a term that is one-sidedly unfair, imposed without real bargaining power, can be void, not just unenforceable on a technicality. A limitation of liability clause is not automatically safe just because it is in writing and both parties signed it.

Red flags

Normal	Red flag	Why it matters
Cap at 12 months' fees or a stated multiple, roughly matched to plausible loss	Cap tied to "fees paid" with no floor, on a low-fee, high-stakes contract (a discounted pilot)	A near-zero cap can mean near-zero recourse for a serious failure
Exclusion limited to "indirect, incidental, consequential and punitive damages"	Exclusion also sweeps in "loss of profit, revenue, business or goodwill, direct or indirect"	Labels ordinary, foreseeable losses as excluded, even ones a court might treat as direct under Sections 73-74
Cap and exclusions apply mutually, on the same terms	Cap protects the vendor's exposure but the customer's own liability was never near that level anyway	A "mutual" clause that is symmetric on paper but one-sided in effect
Carve-outs for fraud, gross negligence, confidentiality breach, and IP infringement	No carve-outs at all, or only fraud carved out	An LoL clause with no carve-outs can cap even a data breach at a token sum
Carved-out claims sit under a separate, higher super-cap (3x-5x fees, or a stated sum)	Carve-outs are technically uncapped with no stated ceiling anywhere	"Uncapped" sounds protective but can mean genuinely unlimited, unplannable exposure
LoL clause states explicitly how it interacts with the indemnity clause	LoL and indemnity are both silent on which one governs if they conflict	Two uncoordinated liability regimes, and an argument fought out after something has gone wrong
Cap applies per claim or per year, matched to contract value	Single lifetime aggregate cap on a multi-year, escalating-fee contract	An early, smaller claim can exhaust the cap, leaving nothing for a later, larger failure

Bad clause → better clause

Bad: "In no event shall either party be liable to the other for any damages arising out of or in connection with this Agreement, whether in contract, tort or otherwise, and each party's total liability shall not exceed the fees paid in the preceding month."

What is wrong: the cap is absurdly low (one month's fees), there are no carve-outs, so even fraud is capped there, and "any damages" excludes recourse almost entirely rather than just consequential loss.

Better: "Subject to the carve-outs below, neither party shall be liable to the other for indirect, incidental, consequential, special or punitive damages, including loss of profit or revenue, arising out of or in connection with this Agreement. Subject to the carve-outs below, each party's total aggregate liability arising out of or in connection with this Agreement shall not exceed an amount equal to the fees paid or payable by Client in the 12 months preceding the event giving rise to the claim. The exclusions and cap in this Clause shall not apply to: (a) either party's indemnity obligations under Clause [X]; (b) breach of confidentiality under Clause [Y]; (c) either party's fraud, gross negligence or wilful misconduct; or (d) Vendor's infringement of Client's intellectual property rights, for which liability shall not exceed [3x fees paid in the preceding 12 months / a stated sum]."

What changed and why: the cap is tied to a 12-month rolling figure rather than one month, the exclusion is limited to indirect and consequential loss rather than "any damages," and the carve-outs form a defined, honest list with its own super-cap rather than none at all, or unlimited exposure.

How it interacts with related clauses

A limitation of liability clause rarely works in isolation. Three clauses decide what it means in practice:

• Indemnity. If your indemnity clause is silent on whether it sits inside or outside the LoL cap, you have two liability regimes never designed to work together. Well-drafted contracts either fold indemnity into the general cap or carve it out with its own super-cap; see our indemnity clause guide on how Sections 124-125 treat indemnity as a distinct promise.
• Insurance. A cap is only worth what the paying party can actually pay. A cyber-liability or professional-indemnity requirement, sized to match the cap, turns the number in the clause into money you can collect.
• Compensation and damages (Sections 73-74). The LoL clause modifies, but does not replace, the statutory baseline it falls back on if struck down or found not to apply.

You can mark up an LoL clause, its carve-outs, and how it lines up against your indemnity clause directly in a document, for free, using Weave, which lets you flag and comment on clauses like this before you send a contract back for negotiation.

US and global contrast

US commercial contracts use similar mechanics, a cap, an indirect/consequential exclusion, carve-outs, but the legal backdrop differs. Many US states enforce even aggressive exclusions, including for a party's own negligence, as long as the language is clear; there is no single federal rule. The "unconscionability" doctrine that can void a one-sided term is applied narrowly, mostly in consumer contracts, not routinely in negotiated B2B deals. In India, Section 23 performs a similar function through a public-policy lens, and its use against ordinary commercial LoL clauses remains the exception, not the rule. Do not assume either country's default applies to the other.

FAQ

What is a limitation of liability clause? It is a clause that caps how much one party can be made to pay the other for a breach, and usually excludes entire categories of loss, like indirect or consequential damages, subject to carve-outs for things like fraud or IP infringement.

Can a limitation of liability clause be struck down under Indian law? Yes, in principle, under Section 23, if it is unconscionable and imposed on a party with materially unequal bargaining power, following the reasoning in Central Inland Water Transport Corporation v Brojo Nath Ganguly (1986 AIR 1571). In practice this is rare between businesses of comparable size, and more likely for standard-form, take-it-or-leave-it terms.

Is a cap of "12 months' fees" reasonable? It depends on the fees relative to the potential harm. On a heavily discounted pilot, or a small monthly fee tied to a mission-critical system, 12 months' fees can be a fraction of realistic exposure. Compare the cap to what a serious failure would actually cost you, not just to convention.

What is the difference between a cap and an exclusion? A cap sets a money ceiling on liability that is otherwise recoverable. An exclusion removes an entire category of loss, commonly indirect, consequential, or punitive damages, from being claimed at all. Most LoL clauses use both together.

Should indemnity obligations be inside or outside the liability cap? There is no single right answer, but silence is the wrong answer. Many contracts carve indemnity out and give it a higher super-cap, because indemnified risks like IP infringement or confidentiality breach are often what parties most want fully covered. Whatever you choose, both clauses should say the same thing.

Does excluding "all liability" actually work? Rarely, and it is a red flag. A clause purporting to exclude all liability, including for fraud, risks being read narrowly by a court, or challenged as unconscionable under Section 23, particularly where one party had no real ability to negotiate it. Honest, specific carve-outs are more enforceable, and more useful to negotiate around, than a clause that overreaches.

This guide gets you to understanding what a limitation of liability clause does and how Indian law treats it. It does not tell you whether a specific cap, exclusion, or carve-out in your contract is enforceable or advisable for your situation, that depends on the specific facts, the bargaining history, and the governing law clause, and is not legal advice. Talk to a lawyer before you rely on, or walk away from, a limitation of liability clause in a live negotiation.