Architects, engineers, surveyors, design and build contractors and construction consultants are routinely asked to sign collateral warranties: short documents, usually presented as standard, often arriving in batches late in a project when refusing feels impossible. Each one extends your legal exposure to a party you never contracted with. Whether your professional indemnity insurance follows that exposure depends on wording, in the warranty and in the policy, and the time to compare the two is before signature.
What a collateral warranty does
Your appointment creates duties to your client. A collateral warranty creates a parallel contractual promise to someone else with an interest in the project: the funder, the purchaser of the completed building, its tenants. Its effect is to give that third party a direct route to sue you for defects in your work, without having to rely on your client to do it. Warranties commonly also allow assignment, so the beneficiary can pass the right on when the building is sold, meaning the party who eventually claims against you may be several transactions removed from anyone you ever met. None of this is inherently unreasonable; it is how funded construction works. It simply means each warranty deserves reading as what it is: a new set of liabilities, lasting years.
The clauses that collide with your insurance
Four recur. First, fitness for purpose. A professional indemnity policy responds to a failure to exercise reasonable skill and care, which is the standard the law normally implies. A warranty that instead promises the works or design will be fit for their intended purpose promises an outcome, not an effort, and liability under such a promise can exist without any negligence at all, which is precisely the liability a professional indemnity policy is commonly not built to cover. Fitness for purpose language belongs on the negotiation list every time it appears.
Second, the obligation to maintain insurance, typically at a stated limit for a fixed number of years after practical completion. This is a real, priceable promise: years of premium at a limit you might not otherwise carry, through market conditions nobody can predict. The usual protective phrase, that cover be maintained provided it remains available at commercially reasonable rates, exists for good reason, and its absence is worth noticing. Third, the required limit and its basis: a warranty demanding a limit for each and every claim is not satisfied by an aggregate policy with the same figure. Fourth, net contribution. Without a net contribution clause, a beneficiary can pursue you for the whole of a loss you contributed to only in part, leaving you to recover the balance from the other culpable parties as best you can. With one, your liability is capped at your fair share. Insurers notice the difference, and so should you.
The order of events that works
The professionals who get this right do the same three things. They agree the form of warranty at appointment stage, when they still have leverage, rather than at practical completion when they have none. They send the draft to their broker before signing, so the promises being made are checked against the policy actually held, the limit, its basis, the treatment of defence costs, and any exclusions that touch the project type. And they keep every signed warranty on a register with its insurance obligations and expiry dates, because an obligation to maintain cover for twelve years is only manageable if somebody remembers it exists. None of this is onerous once it is routine, and it is dramatically cheaper than discovering a mismatch during a claim.
We review appointments and collateral warranties against our clients’ professional indemnity policies as a standard part of arranging cover for construction professionals. Our professional indemnity insurance page explains the underlying cover, and for the related question of non-negligent damage to neighbouring property during works, see our non-negligent liability page.
