When General Liability Coverage Responds: Timing, Operations and Assumed Liability

When General Liability Coverage Responds: Timing, Operations and Assumed Liability

Three Questions Before "Is It Covered?"

A liability policy helps when a business is legally responsible for harm to someone else. Before anyone can say a claim is covered, three separate questions need answers:

• Timing: Which policy period is the right one to look at?
• Classification: Did the harm come from the business's location and active work, or from its sold products and finished work?
• Exclusions: Does a specific rule remove this kind of liability?

Passing one question never guarantees payment. The actual policy and its endorsements (written changes added to the policy) still control every other requirement.

By the end, you should be able to:

• pick the policy period an occurrence policy points to;
• test a claim against a claims-made policy's claim date, retroactive date and reporting deadline;
• classify an injury as premises, ongoing operations, products or completed operations;
• separate liability taken on from another party from a business's own broken promise.

Key words. A policy period is the time a policy is in force, such as January 1 to December 31, 2026. A claim is a demand, such as a letter or lawsuit, asking the insured business to pay for harm. Commercial general liability (CGL) insurance covers a business for bodily injury to other people and damage to their property. A trigger is the event that decides which policy period responds.

Every business, clause and policy below is fictional.

Occurrence Policies: When Did the Harm Happen?

Plain takeaway: An occurrence policy looks at when the injury or damage happened. It does not look at when the careless work was done, when a product was sold or when the claim arrived. Most CGL insurance is written this way.

Worked example

Brightline Electric has occurrence policies for 2025 and 2026. Assume every requirement other than timing is met.

• 2024: Brightline installs faulty wiring in a client's office.
• August 2025: The wiring overheats and damages the office.
• March 2026: The client sends Brightline a claim.

In 2024 there was a mistake, but nothing was damaged yet. In August 2025 the damage happened. That is the event an occurrence policy watches. The 2026 date only tells us when someone asked for money.

Result: The 2025 policy is the timing candidate.

Two tempting mistakes: "The claim arrived in 2026, so use 2026" applies claims-made thinking to an occurrence policy. "The work was done in 2024" treats the construction date as the injury date. Faulty work can sit harmlessly for months before it damages anything.

Try it: occurrence timing

Brightline has occurrence policies for 2025, 2026 and 2027. It installs a water heater in 2025. The heater leaks and damages a client's floor in 2026. The client makes a claim in 2027. Assuming all other requirements are met, which policy is the timing candidate?

Answer: The 2026 policy

The damage happened in 2026. The installation year is when the work was done, and the claim year is when payment was requested. Neither one decides an occurrence policy.

Claims-Made Policies: When Was the Claim First Made?

Plain takeaway: A claims-made policy looks at when a claim is first made against the insured. That must happen during the policy period.

Many claims-made policies also have a retroactive date. Plain takeaway: it is a cutoff in the past. Harm before that date is not covered, even if the claim arrives during the policy period. The retroactive date is not the policy start date. It can be years earlier.

Policies differ on which event must fall on or after the retroactive date. Many professional liability policies use the wrongful act, meaning the mistake itself. Our fictional policy uses the date the injury happened. Always check which event the actual policy names.

Worked example

Fictional Policy B is a claims-made policy for January 1 to December 31, 2026. Its retroactive date is January 1, 2024, and the injury must happen on or after that date. A customer is injured in 2025 and first makes a claim in April 2026. Assume all other requirements are met.

1. Was the claim first made during 2026? Yes, in April.
2. Did the injury happen on or after January 1, 2024? Yes, in 2025.

Result: Policy B is the timing candidate, even though the injury happened before the policy began. The earlier retroactive date reaches back to include it.

Try it: retroactive date

Same Policy B. The injury happened in 2023, and the claim is first made in 2026. Does the claim pass Policy B's timing tests?

Answer: No, it fails the retroactive date test

The claim date fits the 2026 period, but the injury happened before January 1, 2024. Meeting one timing test does not make up for failing the other.

Claims-Made-and-Reported: Two Dates Must Fit

Plain takeaway: A claims-made-and-reported policy requires the claim to be made against the insured and reported to the insurer during the policy period, unless the actual policy gives extra time to report.

Fictional Policy C is claims-made-and-reported for 2026 and gives no extra reporting time. Assume the retroactive date and all other requirements are met.

Case B shows the problem with these policies. A claim that arrives late in the period leaves little time to report it.

Change one fact: Suppose Policy C instead said claims made during the period may be reported up to 45 days after it ends. January 5 is within 45 days of December 31, so Case B would pass the reporting test. Some policies include a window like this and others do not. Never assume one exists.

Try it: two columns

Under the original Policy C, a claim is first made against the insured in October 2025 and reported to the insurer in February 2026. Does it pass Policy C's timing tests?

Answer: No, the claim was first made before the period

The report date falls in 2026, but the claim was first made in 2025. Reporting during the period cannot fix a claim that was first made before the period began.

Where Did the Harm Come From?

Fernway Fixtures sells step stools and installs shelving. Each injury below belongs to a different group.

• Premises: harm connected to a place the business owns, rents or uses. A visitor trips on a torn mat inside Fernway's store.
• Ongoing operations: harm during work still in progress. While Fernway's crew is installing shelves at a client's warehouse, a dropped bracket hits a worker.
• Products: harm from something the business sold, after the business no longer has it and away from its premises. A customer's Fernway stool collapses at home.
• Completed operations: harm caused later by finished work. Months after the warehouse job is done and the client is using the shelves, a shelf gives way.

Policies often group the last two together as the products-completed operations hazard. The test is whether the harm comes from a product that has left the business's hands or from work that is finished. The date a business closed a job file, or stopped operating, does not decide it. A stool that collapses on Fernway's showroom floor is a premises or operations case, because Fernway still has it.

Real policies define "completed" in detail. For example, work can count as completed even if it later needs repair. This lesson uses a simpler fictional rule: work is completed when all contracted work is finished and the customer is using it.

Classification is not a coverage decision. It tells you which part of the policy to read, and endorsements may change how a group is treated.

Try it: classification

The warehouse job is finished, and the client has used the shelves for months. A Fernway technician returns to adjust one shelf. The technician's drill falls and injures a client employee. Which group fits?

Answer: Ongoing operations

The injury came from work happening right then, not from the finished shelves. Finishing the original job does not make every later visit "completed operations." If the finished shelf itself had collapsed, that would point to completed operations.

Contractual Liability: Whose Responsibility Did You Take On?

• Tort: a legal wrong, such as carelessness, that causes harm. Responsibility can exist with no contract at all. If Fernway's crew leaves a cord across a walkway and someone trips, the responsibility comes from the carelessness.
• Hold harmless (indemnity) clause: Plain takeaway: one party agrees to pay for another party's legal responsibility.
• Contractual liability, in insurance: Plain takeaway: taking on another party's liability through a contract. It does not mean every promise in a contract.

Many liability policies exclude contractually assumed liability, then make an exception for an insured contract. Plain takeaway: "insured contract" is a defined list of agreements, and for those agreements the exclusion does not apply. It brings back selected assumed liability. It does not cover every contract obligation.

Worked example

Stagecraft Rigging contracts with Harbor Hall, a concert venue. Fictional clause: "Stagecraft will pay Harbor Hall's legal responsibility for bodily injury to guests arising from Stagecraft's rigging work." Fictional Policy D excludes liability assumed by contract, except in an insured contract, and its schedule names this agreement as an insured contract.

Stipulate that a guest is hurt during the rigging work, sues Harbor Hall, and Harbor Hall demands that Stagecraft pay under the clause.

The only question here is whether the exclusion's exception applies. It does. Stagecraft took on Harbor Hall's liability to an injured guest, in an agreement Policy D names as an insured contract. That does not settle whether the clause is legally enforceable or whether Policy D ultimately pays.

Change one fact: Suppose the contract instead says, "Stagecraft will finish rigging by June 1 or pay $500 per late day." Stagecraft finishes late, and Harbor Hall bills the penalty. Stagecraft did not take on Harbor Hall's liability to anyone. This is Stagecraft's own business promise, so the insured contract exception does not make the penalty covered.

One caution: the legal label on a lawsuit does not settle the question by itself. For example, a claim that a product's breach of warranty caused injury, or damaged property other than the product, is a contract theory. It still cannot be dismissed just because of that label. Other policy terms apply.

Try it: assumed or not?

Stagecraft leases a warehouse, and Policy D names the lease as an insured contract. The lease requires monthly rent. It also says Stagecraft takes on the owner's legal responsibility for injuries to Stagecraft's visitors. The owner demands two things: (1) three months of unpaid rent, and (2) payment of a visitor's injury claim against the owner, from an accident stipulated to fall within that clause. Which demand involves the assumed liability that the insured contract exception addresses?

Answer: Only demand 2

Demand 2 asks Stagecraft to pay the owner's liability to an injured visitor, which is exactly what the lease shifted to Stagecraft. Demand 1 is Stagecraft's own unpaid debt. A contract appearing on the policy's list does not turn every obligation in it into covered liability.

What to Remember

• Treat timing, classification and exclusions as separate questions. Passing one does not mean the claim is paid.
• Occurrence: use the date the injury or damage happened.
• Claims-made: the claim must be first made during the period, and the event the policy names must fall on or after any retroactive date.
• Claims-made-and-reported: check both the claim date and the report date. Count only reporting time the policy actually gives.
• Premises and ongoing operations involve the location and active work. Products and completed operations involve goods that left the business's hands or finished work.
• Contractual liability means taking on someone else's liability, not breaking your own business promise. An insured contract exception brings back only selected assumed liability.