Casualty LinesChapter 522 min read

Product Liability & Recall, when the thing you sold breaks something.

When a manufactured product injures a person or damages property after it has left the manufacturer's hands, that exposure is product liability. When the manufacturer chooses or is forced to pull the product from the market, that is recall. The two are written on different forms, often by different markets, and the line between them is one of the most underappreciated coverage seams in commercial insurance.

§ 01

The mental model

Most product liability is covered under the products and completed operations hazard of a CGL policy. Most recall is covered under a standalone recall or contaminated products policy. The two are adjacent but operationally distinct, and many manufacturers are surprised to learn at the moment of crisis which one their loss falls under.

A defective product hurts someone or damages property. That is a CGL claim, sitting under the products and completed operations hazard. A defective product has not yet hurt anyone but the manufacturer believes it will and pulls it from the market. That is a recall claim, and CGL does not cover it. The cost of pulling product, communicating with distributors and consumers, replacing inventory, and managing the brand impact lives in a separate product recall policy. Manufacturers buying CGL alone for products risk often have no recall coverage. Manufacturers who only buy recall have no third-party liability protection. The two coverages are complementary and neither substitutes for the other.

Anchor concept

Bodily injury or property damage caused by a product is CGL territory. Pulling product before injury is recall territory. The trigger that distinguishes them is whether harm has occurred. Once you internalize that, the rest of the line organizes itself around it.

§ 02

Products vs completed operations

The CGL form bundles "products" and "completed operations" together as a single hazard for a reason: they are structurally similar but apply to different kinds of business. Both respond to losses that occur after the insured's work has been completed and the product or service has left the insured's control.

Products hazard

Bodily injury or property damage arising out of the insured's product after it has been sold, distributed, or otherwise leaves the insured's possession. The defining characteristic: the product is a physical thing, and the loss occurs away from the insured's premises (typically at the customer's location).

Completed operations hazard

Bodily injury or property damage arising out of the insured's work after the work has been completed. The defining characteristic: the loss occurs because of work the insured performed, not because of a product the insured sold. A roofer's work that fails three years later is completed operations. A manufactured shingle that fails three years later is products.

Why they share a single hazard

Both share the long-tail discovery problem: the loss may not surface until years after the policy expires. Both share a common rating approach (sales for products, payroll or contracts for completed operations). Both share common exclusions, particularly for the insured's own product or work (the "your product" and "your work" exclusions, which carve out the cost of replacing or repairing the defective item itself).

§ 03

Manufacturing, design, warning

Product liability claims fall into three legal theories, and the underwriter pricing the risk needs to think about all three.

Manufacturing defect
The specific unit that caused the loss was defective compared to the design. A bottle that exploded because of a contamination issue specific to that bottle. Quality control failure. Often easier to defend because the issue is unit-specific.
Design defect
The product as designed was unreasonably dangerous, even if manufactured correctly. The fuel tank that ruptures in a side impact because of where it sits in the vehicle. Hits every unit of the product, not just one. The most expensive theory because it can support a class action.
Failure to warn
The product was reasonably safe if used as intended, but the manufacturer failed to warn of risks the user could not be expected to discover. The product is not defective; the documentation is. Often the most argued theory in pharmaceutical and chemical product cases.

The three theories are not mutually exclusive. A single claim often pleads all three to maximize recovery options. The defense burden, the expert witness costs, and the discovery scope are different for each, and the underwriter pricing a risk should think about which theories are most plausible given the product type.

§ 04

The vendor's endorsement

A retailer selling a product manufactured by someone else has product liability exposure even though they did not make the product. State strict liability laws often allow injured parties to sue any party in the chain of distribution, including downstream vendors. The vendor's endorsement to the manufacturer's CGL policy extends coverage to vendors who carry the product.

How it works

The manufacturer's CGL policy is endorsed to add a defined vendor as an additional insured for product liability arising out of the manufacturer's product. The vendor gets coverage on the manufacturer's policy at the manufacturer's limit, without having to carry their own product liability insurance for that line of products. In exchange, the vendor often pushes the manufacturer to carry larger limits and to name them on a continuing basis through the supply contract.

The exclusions inside the endorsement

Standard vendor's endorsements exclude coverage for the vendor's own products, the vendor's repackaging or relabeling, the vendor's failure to maintain the product, and similar carve-outs that target risks the vendor created themselves. The manufacturer is only on the hook for losses arising out of their product as it left their facility, not for what the vendor did to it afterward.

Where vendor endorsements break

The vendor's endorsement is one of the most-litigated coverage provisions in commercial insurance. Three common breakdowns: (1) the endorsement names the wrong vendor entity (a parent instead of an operating subsidiary); (2) the vendor does something to the product (relabeling, repackaging, modifying) that triggers an exclusion the vendor did not understand; (3) the manufacturer's policy has been exhausted by other claims, leaving the vendor without the protection they thought they had.

§ 05

Recall as a separate trigger

Product recall coverage responds when a manufacturer has to remove or correct a product that has not necessarily caused injury but is suspected of being likely to cause injury or has been found to fail to meet some required standard. The trigger is the recall event itself, not bodily injury or property damage.

Three categories of recall

  • Voluntary recall. The manufacturer decides on its own initiative to remove the product, often after internal testing or initial complaints suggest a problem.
  • Government-ordered recall. A regulatory body (FDA, USDA, NHTSA, CPSC, etc.) directs the manufacturer to recall the product. Mandatory.
  • Customer-requested recall. A major customer (a retailer, a distributor, a downstream manufacturer) demands the recall as a condition of continuing the relationship. Often de facto mandatory even though it is not a regulatory order.

The covered costs

Recall coverage typically includes notification expense (telling customers, retailers, the public), retrieval expense (logistics of getting the product back), destruction or refurbishment cost, lost gross profit during the recall period, transportation, storage, and consultant or crisis management fees. Higher-end products extend to brand rehabilitation expense and increased advertising. Some products extend to extortion (if the recall was triggered by a tampering threat).

§ 06

First-party recall expense

The structural distinction between CGL products coverage and recall coverage is third-party vs first-party. CGL responds to claims made against the manufacturer by injured parties (third-party). Recall responds to costs incurred by the manufacturer itself (first-party).

Why this matters

A manufacturer pulling a product from the market is incurring its own expense to prevent third-party injury. Without recall coverage, those expenses are uninsured. The manufacturer is paying out of pocket to prevent a CGL claim, which the CGL policy benefits from but does not reimburse. Recall coverage was created precisely to address this gap.

Common products with heavy recall exposure

  • Food and beverage (most frequent recalls, regulated by FDA / USDA)
  • Pharmaceuticals (severe consequences, FDA oversight)
  • Automotive parts (NHTSA, large fleet recall logistics)
  • Consumer products (CPSC, broad distribution channels)
  • Children's products (heightened regulatory environment)
  • Electronics with safety implications (battery fires, etc.)
  • Medical devices (FDA, often class II / III device tracking requirements)
§ 07

Contractual indemnity flow-down

Most product manufacturers operate under master supply agreements with their customers. These agreements include indemnity provisions that flow product liability and recall obligations down the supply chain. Understanding how those indemnity obligations interact with insurance is one of the hardest parts of underwriting product risk.

Hold harmless and indemnification

A typical supply agreement requires the manufacturer to indemnify the customer for losses arising out of the product. Insurance picks up that contractual obligation through the contractual liability provisions of the CGL form. The CGL covers liability the insured has assumed under an "insured contract," which includes most arms-length supply agreements.

Where contractual exposure exceeds insurance

Some indemnity provisions go beyond what insurance will cover. Indemnity for the customer's own negligence, indemnity for losses without limit even where insurance has a sub-limit, indemnity for consequential damages explicitly excluded by the policy. The manufacturer signs the contract anyway because the customer demands it. The result: contractual exposure that insurance will not respond to. This is a common gap that surfaces only when a claim happens.

Underwriters increasingly request sample customer contracts during the application process. The indemnity language drives more of the actual exposure than the application's narrative description of the business.

§ 08

Underwriting and rating

Product liability rating starts with sales as the exposure base, with rates that vary enormously by product class. A small machine shop's product premium per $1,000 of sales is a fraction of a fireworks manufacturer's premium per $1,000 of sales.

Key underwriting factors

  • Product classification. The ISO classification code that captures product type and inherent risk. Drives the base rate.
  • Sales mix. Different product lines often have very different rates. The underwriter wants the breakdown.
  • Quality control program. ISO 9001 or equivalent certification, internal QA processes, supplier qualification, batch traceability, retention samples.
  • Distribution channel. Direct to consumer, retail, OEM, industrial. Different channels carry different risk profiles.
  • Geographic distribution. US-only vs international. Some jurisdictions are much more product-friendly to plaintiffs.
  • Loss history. Five to ten years of product claims and recalls. Patterns matter more than any single loss.
  • Contractual obligations. Sample customer contracts, indemnity language, insurance-and-indemnity certificate requirements.
  • Recall history. Prior recalls, even if uninsured, are a leading indicator of future loss.

Capacity considerations

Primary product liability typically caps in the $1M-$5M range per occurrence. Excess and umbrella towers extend above. For high-risk products (firearms, sporting goods, heavy equipment, pharmaceuticals), capacity may be constrained and pricing reflects scarcity. Recall capacity is generally lower than CGL capacity, with $5M-$25M towers typical for mid-sized manufacturers and larger for major consumer brands.

§ 09

Where IDP earns its keep

Product liability submissions are document-heavy across two dimensions: the application itself, and the supporting documentation about the products and the business. Sample contracts, distribution agreements, quality control documentation, regulatory filings, recall histories, loss runs, and product literature all show up in a typical submission.

01
Intake
Application, supps, contracts, recall history.
02
Classify
Product class, distribution channel, geography.
03
Extract
Sales mix, contractual indemnity, prior recalls.
04
Validate
Application vs contract consistency, QA evidence.
05
Triage
Class hazard, severity profile, capacity fit.
06
Underwriter
Pre-populated workspace + indemnity flag.

Indico use cases

  • Indemnity clause extraction. Pull hold-harmless and indemnification language from sample customer contracts and surface terms that exceed standard insurance coverage.
  • Recall history normalization. Convert FDA, CPSC, and NHTSA recall reports into structured records with severity flags and pattern analysis.
  • Product mix extraction. Parse sales schedules and product catalogs to derive the rating mix the underwriter needs.
  • Quality control document classification. Tag uploaded supporting documents (ISO certifications, audit reports, supplier qualifications) so the underwriter can find what they need without hunting.
  • Loss run analysis. Identify product-specific patterns (single product driving frequency, geography concentration, severity development).
Where the demo lands

For a product underwriter, the contract and indemnity extraction agent is the highest-leverage demo. Five sample customer contracts in, the agent surfaces every indemnity provision, flags those with broader-than-insurance scope, and produces a one-page risk summary the underwriter uses to price the contractual exposure. The work is mechanical at scale and the consequences of missing it are large.

Chapter 5 · Casualty Lines · 22 min read

Product Liability & Recall — Cheat Sheet

When a manufactured product injures a person or damages property after it has left the manufacturer's hands, that exposure is product liability. When the manufacturer chooses or is forced to pull the product from the market, that is recall. The two are written on different forms, often by different markets, and the line between them is one of the most underappreciated coverage seams in commercial insurance.

The mental model: Bodily injury or property damage caused by a product is CGL territory. Pulling product before injury is recall territory. The trigger that distinguishes them is whether harm has occurred. Once you internalize that, the rest of the line organizes itself around it.

Watch for

Key terms

Products hazard · CGL coverage trigger
Completed ops · Work-after-completion
Vendor's endorsement · AI for retailers
Insured contract · CGL contractual liability
FDA / CPSC / NHTSA · Recall regulators
First-party recall · Manufacturer's own cost

If you remember three things

Product liability is third-party and lives on CGL. Recall is first-party and needs its own policy. The vendor's endorsement extends manufacturer's coverage to retailers but only for losses arising from the manufacturer's product as delivered.