Duty to Defend vs. Duty to Indemnify: The Two-Word Difference That Costs Six Figures

Every subcontractor has fought about indemnity. It's the clause your insurance broker warns you about, the one anti-indemnity statutes in most states exist to police, the one leadership actually reads before signing. But inside the standard formulation — 'Subcontractor shall defend, indemnify, and hold harmless…' — the first verb does more damage than the other two combined, and it is the one that almost never gets negotiated.

The duty to indemnify is a promise about the end of a claim: if there's a judgment or settlement within the clause's scope, you reimburse it. The duty to defend is a promise about the beginning: when a claim is filed, you pay for the defense — attorneys, experts, discovery — as the bills arrive. Those bills do not wait to find out whether you were at fault. In commercial construction litigation, where defense costs routinely run into six figures before any decision on the merits, the defense obligation is frequently worth more than the indemnity it accompanies.

In most jurisdictions, the duty to defend is triggered by the allegations of the complaint, not by the facts as they are eventually found. If the owner's lawsuit alleges that your work contributed to the loss — even wrongly, even implausibly — a defense obligation keyed to claims 'arising out of' your work can be triggered by the pleading alone. You can win the case completely and still have funded years of someone else's litigation.

California made the point brutally clear in Crawford v. Weather Shield Mfg., Inc. (2008), where the state Supreme Court held that a subcontractor's contractual duty to defend arose immediately upon tender of the claim — and survived even though the subcontractor was ultimately found not negligent at trial. The lesson generalizes: unless your contract expressly ties the defense obligation to a finding of fault, courts in many states will read it as an immediate, allegations-based duty. That is the default you are silently accepting every time the clause goes through unmarked.

Subcontractors often assume their state's anti-indemnity statute protects them from the worst of this. Sometimes it does — several statutes expressly reach defense obligations. But in a meaningful number of states, the statute voids indemnity for another party's negligence while saying nothing clear about defense costs, and GC-drafted contracts are written to live in exactly that gap: the indemnity is trimmed to what the statute allows, while the defense obligation remains broad, immediate, and allegations-triggered.

This is also where the interplay with insurance gets dangerous. Your CGL policy's duty to defend runs from your insurer to you — it does not automatically cover a contractual promise to defend someone else. Contractual-liability coverage may pick up an assumed defense obligation, but the scope depends on the policy, the endorsements, and the state. The safe assumption is the uncomfortable one: an unmarked 'defend' can be an uninsured, out-of-pocket obligation. Pair this article with our review of additional-insured endorsements and waivers of subrogation — the three provisions travel together and should be negotiated together.

Picture a $400,000 mechanical subcontract on a project where the owner later sues the GC for water intrusion, naming every trade that touched the building envelope. The complaint alleges — generically, as complaints do — that each subcontractor's work 'contributed to' the defects. Under a broad defense clause, the GC tenders its defense to all of them. Your share of joint defense costs starts accruing immediately: document collection, depositions, expert reports. Two years later, the forensic analysis clears your scope entirely. Under an allegations-based defense duty with no fault trigger and no reallocation language, the money you spent defending the GC may simply be gone.

Now scale that across a portfolio. The 2025 Arcadis data putting average U.S. dispute values at $60.1 million describes the headline cases, but the defense-cost dynamic operates identically on ordinary disputes — it just destroys smaller companies faster. This is a classic example of what we call the hidden cost of rubber-stamping GC contracts: the clause costs nothing on signing day and compounds silently until the day it costs everything.

First choice: strike 'defend' and convert the clause to indemnity-plus-reimbursement — you will reimburse reasonable defense costs attributable to your adjudicated share of fault, after fault is determined. GCs resist, but it's a principled position, and in states whose anti-indemnity statutes reach defense obligations it may simply be the law. Second choice: keep the defense duty but tie it expressly 'to the extent' claims arise from your negligent acts or omissions, with reallocation once fault is decided — so you fund your share of the fight, not the whole war.

Third choice, if you must accept a broad duty: cap it, coordinate it with your insurance program (confirm contractual-liability coverage actually reaches it), and demand control-of-defense and consent-to-settle rights so you're not writing blank checks to lawyers you didn't choose. And in every scenario: tender every claim to your own insurer immediately — notice deadlines destroy more coverage than exclusions do. An automated first pass helps here precisely because this clause hides in plain sight: it looks like the indemnity everyone already argued about, and it isn't.

One: does the clause say 'defend'? If so, is the duty triggered by allegations or by an actual finding of fault? Two: does it cover claims 'arising out of' your work (broad) or only 'to the extent caused by' your negligence (narrow)? Those four words move six figures. Three: who picks the lawyers, and do you have consent rights over any settlement paid with your money? Four: does your contractual-liability insurance actually respond to an assumed defense obligation in this state — have your broker confirm in writing, not by phone.

Five: is there reallocation language — a mechanism to true-up defense costs after fault is decided — or does the money simply stay where it landed? A clause that fails all five is not a dealbreaker on every project, but it should be a priced risk, not an unread one. If the job is thin-margin and the counterparty litigious, the honest answer may be the one we wrote about in why contractors walk away from profitable-looking work.

Read the verbs. 'Indemnify' is a promise about losses; 'defend' is a promise about legal bills that start immediately, turn on allegations rather than facts, may survive your complete vindication, and may sit outside both your state's anti-indemnity statute and your insurance program. It deserves at least the negotiating attention the indemnity itself gets — and in a defense-cost world, usually more.

None of this is legal advice, and defense-and-indemnity law is aggressively state-specific — the difference between Crawford-style immediate duties and fault-based reimbursement regimes is exactly the kind of question that deserves an hour of qualified counsel's time on your contract template. What a good first-pass review buys you is knowing the clause is there, which flavor it is, and what it could cost — before you've signed it, not after the tender letter arrives.