Indemnity
By Gareth HoyleLast updated How we research
An indemnity is a contractual promise by one party to cover the other's losses if a named risk occurs, such as a third-party claim over licensed data.
An indemnity is a promise in a contract that one party will compensate the other for specified losses, usually those caused by third-party claims. In a data license the buyer typically asks the seller to indemnify it against claims that the data infringed someone's rights or breached privacy law.
How are indemnities limited in data deals?
Indemnities in data deals are usually limited by caps, time limits and exclusions. Nixon Peabody's October 2026 alert on buying AI training data names representations, indemnities, liability caps, holdbacks and insurance as the terms that decide who bears the risk.
| Term | What it does |
|---|---|
| Warranty | A statement of fact about the data, such as who owns it |
| Indemnity | A promise to cover losses from a named risk |
| Liability cap | A maximum amount payable, often tied to the fee |
| Holdback | Part of the price held back for a set period to meet claims |
Because a trained model cannot give data back, an uncapped indemnity can outlast the deal by years. Take independent legal advice before you agree one, and see how perpetual licenses change the exposure. The licensing agreement checklist and chain of title entries cover the related clauses.