Guarantee vs Indemnity: The Difference Explained (2026)

The short answer A guarantee is a promise to answer for someone else's default. An indemnity is a promise to cover someone else's loss. They look similar because both shift financial risk onto a third party, but they work differently in law, and the difference decides how easily you can actually recover money. This page explains each plainly, then sets out the practical differences. It is general information, not legal advice, so confirm the position for your jurisdiction and your specific wording with qualified counsel. ## What a guarantee is A guarantee is a secondary obligation. The guarantor promises that if the main party (the principal debtor) fails to perform or pay, the guarantor will step in. A classic example is a parent company guaranteeing a subsidiary's loan, or a director personally guaranteeing a company's lease. The key word is secondary. A guarantor's liability depends on, and mirrors, the main party's liability. If the underlying debt does not exist, is unenforceable, or is reduced, the guarantee usually follows. If the main contract is varied without the guarantor's consent, the guarantor may be released entirely. Because a guarantee is a secondary promise about someone else's debt, in common-law systems it typically must be in writing and signed to be enforceable. ## What an indemnity is An indemnity is a primary obligation. The indemnifier promises to compensate the other party for a defined loss, full stop, regardless of whether anyone else has defaulted. For example, a supplier might indemnify a customer against losses caused by a product defect, or a seller might indemnify a buyer against a specified tax liability. The key word is primary. The indemnifier's duty does not depend on a third party's default or on the validity of some other obligation. If the defined trigger event happens and the loss follows, the indemnity responds on its own terms. That independence is what makes an indemnity powerful for the party protected by it. ## The practical differences Primary versus secondary. This is the heart of it. A guarantee only bites when the main party defaults and tracks that party's liability. An indemnity stands on its own and responds to the defined loss directly, without needing to prove a third party's default. What survives a problem with the main contract. If the underlying obligation is void, discharged, or changed without consent, a guarantee can fall away with it or be released. A properly drafted indemnity can survive, because it is not parasitic on the main obligation. This is why a well-advised party often asks for an indemnity as well as, or instead of, a guarantee. Formality. In common-law jurisdictions a guarantee generally must be evidenced in writing and signed to be enforceable, because it is a promise to answer for another's debt. An indemnity, being a primary promise, does not always carry the same statutory writing requirement, though it should always be written clearly. How you recover. Under a guarantee, you usually pursue the main party and then the guarantor, and your recovery is shaped by the main party's position and by the ordinary rules on proving loss. Under an indemnity, you claim directly for the defined loss on the indemnity's own terms, which can allow a wider or more certain recovery depending on the wording. Who bears the risk. Both shift risk, but an indemnity places it more firmly and more independently on the indemnifier, which is why indemnifiers negotiate hard to cap and qualify them. A guarantor's risk is bounded by the main party's liability. ## When each is used A guarantee is the natural tool when you want a stronger party to stand behind a weaker one's existing obligation: a parent behind a subsidiary, a director behind a company, a group behind a trading entity. The protected party accepts that recovery follows the main debt. An indemnity is the natural tool when you want certain, direct protection against a specific risk, whatever its cause: a known tax exposure, a particular compliance liability, intellectual-property infringement, or losses from a defined event. Commercial contracts frequently use indemnities to allocate named risks to the party best able to control or absorb them. Many real-world contracts combine both. A lender may take a guarantee from a parent and, in the same document, an indemnity so that if the guarantee fails for a technical reason, the indemnity still responds. Because the two are often drafted side by side and even in a single clause, the exact words decide which obligation you actually have, and courts look at substance, not the label. ## A note on caps, limits and fairness Indemnities in particular are often negotiated down: capped in amount, limited in time, narrowed to direct losses, and made subject to the other party mitigating its loss. A guarantor will similarly want protections, such as a requirement to be told of defaults and a release if terms change materially. None of these labels is a magic word, so the enforceability and reach of either promise comes from careful drafting and from the law of the governing jurisdiction. ## Keeping these clauses under control Guarantees and indemnities are where quiet, long-tail risk hides in a contract, and the danger is rarely one clause read in isolation. It is losing track of which agreements carry which promises, whether an indemnity is capped, and whether a variation has quietly released a guarantor. Drafting the clause, reviewing it for exposure, tracking where it sits across a portfolio, and keeping its version history are all the same body of work, and far easier when they live together. That single-workspace idea is what Adira is built around for business contracts: draft, understand, edit, track and govern an agreement in one place, with per-clause risk visible and built on frontier AI. Whatever you use, the point stands: a guarantee answers for another's default, an indemnity answers for a defined loss, and the words you sign decide which you really have.