The indemnity clause is usually the densest page in a commercial contract, and often the page the other side hopes you will skim. It appears in supplier agreements, commercial leases, construction contracts and services agreements, and it decides who carries the cost when something goes wrong. If you are about to sign a draft that contains one, or you are writing terms for your own business, the clause repays close attention before it becomes binding.
What an indemnity clause does is simpler than the wording makes it look. It is a contractual promise by one party to compensate the other for specified loss or damage. The High Court has described an indemnity as a promise to keep the promisee harmless against loss resulting from a transaction with a third party: Sunbird Plaza Pty Ltd v Maloney (1988) 166 CLR 245. It is a risk allocation device. The parties decide up front which of them is best placed to bear each category of risk, and the clause turns that decision into a direct obligation to pay, without the need to prove a breach of some other term of the contract.
An indemnity is also different from a guarantee. A guarantor answers for another party's default, so the guarantee is a secondary obligation that responds only when the primary debtor fails to pay. An indemnifier must pay as soon as the covered loss occurs, regardless of whether anyone else has defaulted. The Federal Court has noted the two are frequently confused, and instruments that say "guarantee and indemnity" can behave differently depending on how they are drafted: Re Taylor; Ex parte Century 21 Real Estate Corporation [1995] FCA 435.
Anatomy of an indemnity clause
Most indemnity clauses answer four questions: who is protected, what is covered, when the obligation is triggered, and how the payment obligation works. The drafting choices around each question decide how much risk actually moves from one party to the other.
Who is protected and who gives the promise
An indemnity names the party giving the promise (the indemnifier) and the party protected (the indemnified party). Two drafting choices matter here:
- Protected class: drafts prepared for the indemnified party commonly extend cover to "the company, its related bodies corporate, officers, employees and agents". Each additional name widens the exposure, so if you are the indemnifier, insist the protected class is confined to the contracting party. If you are the indemnified party, make sure everyone who actually does the work sits inside the clause, or a claim by a subcontractor or employee can fall outside it.
- Mutual indemnities: in services and joint venture agreements, each party often indemnifies the other for losses caused by its own breach or negligence. A mutual indemnity is usually fairer than a one-way clause, but it still needs the same scrutiny of scope, because each side is both giving and receiving.
What losses are covered
The heart of the clause is the definition of loss. A comprehensive definition sweeps in several categories:
- Tangible losses: damage to property, personal injury and death.
- Intangible losses: misuse of confidential information and infringement of intellectual property rights.
- Pure economic losses: lost profits, the cost of rectifying defective work, and statutory fines or penalties.
- Third-party claims: liability to customers, subcontractors or members of the public arising from the contract's activities.
Two traps are worth knowing before you negotiate. First, a broad definition of loss can cover amounts that would be too remote to recover as damages for breach of contract, because the remoteness rules that limit damages do not automatically apply to an indemnity. The clause can therefore create liability that would not otherwise exist at law. Second, an indemnity that expressly covers the statutory penalties or fines of the indemnified party may not be enforceable, and the Commonwealth's own guidance proceeds on the basis that an indemnity should not meet the civil or criminal penalties of the indemnified person. If penalties are intended to be covered, the point should be negotiated and made express rather than assumed.
When the indemnity is triggered
The trigger describes the events or circumstances that activate the obligation. Trigger language usually connects the loss to the contract's activities through words such as "arising out of", "in connection with", "caused by" or "as a result of". The wider the connector, the wider the exposure:
- "Arising out of" and "in connection with": these connectors are wide. They catch losses that merely relate to the activity, whether or not the loss was a direct consequence of it.
- "Caused by": this connector is narrower. It imports a causal link between the activity and the loss, which is usually harder for the claimant to establish than a mere connection.
- "Liable": the word implies a legal obligation to pay. An indemnity that covers "liability" only responds where the indemnified party is actually legally obliged to pay a third party, not where it has volunteered a payment.
The Commonwealth's guidance makes the same point about specificity: an indemnity for "damage to property" exposes the giver to any property damage at all, whereas an indemnity confined to damage caused by the giver's own use of the property is a much smaller promise: Resource Management Guide 414 (the RMG 414).
How broad the trigger is: fault and no-fault indemnities
This is usually the main battleground in negotiation. A no-fault indemnity is triggered whether or not the indemnified party was at fault. A fault-based indemnity only responds where the loss arises out of, for example, "the negligence of the indemnifier", which requires the claimant to prove actionable negligence first:
- No-fault example: a lease clause under which the tenant indemnifies the landlord against all loss "arising from the tenant's use and occupation of the premises" responds to any loss connected with that use, regardless of fault.
- Fault-based example: an indemnity against loss "arising out of the negligence of the contractor" requires the principal to prove negligence before the indemnity responds.
Whether the indemnified party's own negligence is covered is the point most worth resolving in words. Courts construe an indemnity according to its wording and the contract as a whole, and decisions on other indemnities in other contracts are of limited assistance: Erect Safe Scaffolding (Aust) Pty Ltd v Sutton (2008) 72 NSWLR 1. Neither side should rely on an implication that its own negligence is or is not covered. The indemnified party wants an express carve-in for its own negligence; the indemnifier wants an express carve-out, plus an exclusion for fraudulent, wilful or reckless acts of the indemnified party, which is the approach the Commonwealth's guidance takes.
Hold harmless or make good: how the payment obligation works
The drafting of the payment obligation matters more than most people realise, because it can change when the limitation period starts running. Two common formulations produce different results:
- Hold harmless: the indemnifier is in breach of the contract as soon as the indemnified party suffers the loss or damage. The limitation period starts running from the date of the loss.
- Make good: the indemnifier is only obliged to act when called upon to make good the loss. It is not in breach until a claim is made and refused, so the limitation period starts running from that later point.
The High Court analysed this distinction in Sunbird Plaza Pty Ltd v Maloney (1988) 166 CLR 245, where it described the "hold harmless" promise as one that puts the indemnifier in breach as soon as loss is suffered. Because the choice affects when claims must be brought, parties often remove the uncertainty by adding express time limits, for example that claims must be made during the term of the contract or within a stated period after it ends, rather than leaving the indemnity open-ended.
What the indemnifier gets in return
An indemnity that gives the indemnified party money but gives the indemnifier nothing is a one-way street. Well-drafted clauses build in protections for the party giving the promise:
- Notification: the indemnified party must notify the indemnifier promptly of any claim or potential claim.
- Control of defence: the indemnifier gets the right to conduct or participate in the defence of claims against the indemnified party, and to take over any litigation related to the indemnity.
- Subrogation: after paying out, the indemnifier steps into the indemnified party's rights against third parties who caused the loss, so part of the cost can be recovered back.
- Mitigation: the indemnified party must take reasonable steps to minimise the loss, and the indemnifier only pays for liabilities reasonably incurred.
- Carve-outs: fraud, wilful misconduct and reckless acts of the indemnified party fall outside the cover.
These controls are standard in institutional indemnities. RMG 414, for instance, describes subrogation and notification clauses that give the Commonwealth the right to conduct the defence and require full assistance from the indemnified party, and directs that cover not extend to malicious, fraudulent, wilful or reckless acts.
How long it runs and how much it can cost
Two numbers decide the size of the exposure: how long the indemnity survives, and whether the amount is capped.
- Time limits: the clause can limit cover to claims made during the term, or during a stated survival period after the contract ends. Without a limit, the indemnity runs open-ended, which the Commonwealth's guidance expressly warns against.
- Caps: a maximum financial limit, expressed per claim and in aggregate, converts an open-ended promise into a managed risk. RMG 414 describes the express inclusion of a maximum financial limit as an important mechanism to protect the position of the party giving the indemnity.
- The cap question: whether the cap also covers third-party claims, and whether it applies per event or across the whole contract, is often the last detail settled in negotiation, and the one that gets overlooked.
Optional clauses worth asking for
Depending on the transaction, an indemnity clause is often accompanied by one or more of these provisions:
- Insurance back-to-back: a requirement that the indemnifier maintain specified insurance for the life of the indemnity. The trap is that liability policies often exclude cover for contractually assumed liability, so the indemnity and the insurance policy need to be checked together or you can be left with an uninsured gap.
- Parent company guarantee: where the indemnifier is a special purpose entity with few assets, a guarantee from its parent makes the promise worth something.
- Sole remedy: a clause making the indemnity the exclusive remedy for the covered losses, cutting off separate damages claims for the same loss.
- Termination right: the right to end the indemnity when it is no longer needed, rather than letting it run with the contract.
- Claims-made limitation: a provision that only claims made during the term, or within a stated window, are covered.
If the other party is a Commonwealth agency, a different framework applies. Only the Finance Minister, or a delegate, can grant an indemnity on behalf of the Commonwealth under s 60 of the Public Governance, Performance and Accountability Act 2013 (Cth), and the accountable authority must govern the entity in a way that promotes the proper use of public resources under s 15. RMG 414 sets out the considerations that must be weighed before an indemnity is granted, including time limits, the use of commercial insurance, reserving a termination right, maximum financial limits, and subrogation and notification clauses. If you are negotiating with an agency, expect its draft to reflect that framework and to resist open-ended promises.
How an Artificer Legal lawyer would review your indemnity clause
The value a lawyer adds to an indemnity clause is mostly in the questions asked before drafting starts. An Artificer Legal practitioner would begin by mapping who bears each risk under the common law and statute if the contract had no indemnity at all, then test whether the clause improves or worsens your position against that baseline. The review then works through the clause in order: the protected class, the definition of loss, the trigger words, the caps and time limits, and finally the controls of notification, subrogation and insurance. We would push back on "arising out of" language where you are the indemnifier, and resist caps and carve-outs where you are the indemnified party. We would check that the indemnity and your insurance line up so there is no uninsured gap, and confirm the clause is enforceable, particularly where it touches penalties, deliberate wrongdoing or third-party liability. The negotiation order matters too: scope first, then quantum, then process.
Why the trigger words decide who pays
The drafting choice that most often separates an indemnity that works from one that does not is the trigger language. The difference between "all loss arising out of the tenant's use and occupation" and "loss caused by the negligence of the tenant" is the difference between a promise that responds to almost any loss and one that requires proof of fault first. Everything else, the caps, the time limits and the carve-outs, operates on top of that choice. If the trigger is drawn wider than the risk you actually intend to transfer, no amount of careful negotiation elsewhere will contain it.
For most businesses, the practical process is straightforward. Before you sign anything containing an indemnity, identify the real risks of the transaction, decide which party is best placed to bear each one, and then read the clause against that list. The clause should name who is protected, define the losses covered, choose trigger words that match the risk you intended to move, set the time limits and caps, and give the indemnifier the controls and carve-outs that keep the promise honest. If any of those elements is unclear, or the clause is one you did not draft, that is the point at which a commercial lawyer earns their fee.