+150 XP

Manufacturing · Product liability and safety recalls: who pays when a product hurts someone

In 2014, Takata Corporation's airbag inflators started killing people. A chemical propellant inside the inflator could degrade with heat and humidity, causing the metal canister to rupture and spray shrapnel into the passenger cabin. By the time the dust settled, over 100 million inflators had been recalled worldwide, at least 30 deaths were linked to the defect, and Takata went bankrupt. Honda, the automaker that used the most Takata inflators, paid billions in recall costs, settlements, and regulatory fines.

The Takata case is the clearest modern illustration of a question every manufacturing executive eventually faces: when a product hurts someone, who pays, and who decides that it must be pulled from the market? The answer runs through three separate systems that interact but do not always agree: a federal safety regulator, civil courts, and mandatory reporting law. Understanding how they fit together is not optional literacy for anyone in manufacturing, supply chain, or product design.

The regulator: CPSC and its sector cousins

The Consumer Product Safety Commission (CPSC) is the primary US federal agency overseeing safety of consumer products, from toasters to toys to furniture. It was created by the Consumer Product Safety Act of 1972 (CPSA). The CPSC can set mandatory safety standards, ban dangerous products, and order recalls.

Manufacturing is not governed by one safety regulator; it is governed by several, split by product category:

  • CPSC: general consumer products
  • NHTSA (National Highway Traffic Safety Administration): motor vehicles and equipment, including the Takata inflators
  • FDA (Food and Drug Administration): medical devices, drugs, food packaging
  • OSHA (Occupational Safety and Health Administration): worker safety inside the plant, not the product itself

The EU equivalent structure runs through the General Product Safety Regulation (GPSR), which took full effect in December 2024, and sector-specific frameworks like the Medical Device Regulation (MDR). The EU also runs a rapid alert system called Safety Gate for dangerous non-food products across member states. See the European Commission's Safety Gate portal for live recall data.

These regulators do not typically pre-approve products before sale (medical devices and drugs are the major exception). Most manufacturing operates on a post-market surveillance model: sell first, monitor for harm, recall if needed. That model is exactly what creates the speed-versus-exposure trade-off in this lesson's hook.

Mandatory reporting: the clock starts before the public knows

Here is the detail that trips up many operators: manufacturers do not get to decide quietly whether a defect is "serious enough" to report.

Under the CPSA, manufacturers, importers, distributors, and retailers must report to the CPSC within 24 hours of obtaining information that reasonably supports the conclusion that a product contains a defect that could create a substantial risk of injury, or that it violates a safety standard. This is Section 15(b) of the Act. Failing to report in time is itself a legal violation, separate from the underlying defect.

NHTSA has a parallel rule for vehicles: manufacturers must report defects and open investigations, and have historically been required to notify NHTSA within 5 business days of determining a safety defect exists under the TREAD Act (2000), passed after the Ford/Firestone tire failures.

This is where Takata went wrong twice. Internal testing had shown inflator ruptures years before the company disclosed the problem publicly. Delayed reporting turned a product defect into a separate legal and criminal matter: Takata pleaded guilty to wire fraud in 2017 and paid a $1 billion criminal penalty, on top of recall costs.

Practical implication: legal and quality teams inside a manufacturer are on a countdown the moment a defect signal appears internally, whether from field complaints, warranty claims, or lab testing. Speed to disclosure is now a compliance obligation, not a PR choice.

Tort liability: the parallel track courts run

Even when a regulator says a product is fine, a manufacturer can still be sued and lose. This is the civil liability track, governed mostly by state law in the US (there is no single federal product liability statute).

Three legal theories dominate:

  1. Design defect: the product's design itself is unreasonably dangerous, even if manufactured perfectly (the Takata ammonium nitrate propellant was ultimately treated as a design defect).
  2. Manufacturing defect: a specific unit deviated from its intended design, e.g., a contaminated batch.
  3. Failure to warn: the product is inherently risky in some use case and the manufacturer did not adequately disclose that risk.

Most US states apply strict liability: a plaintiff does not need to prove the manufacturer was negligent, only that the product was defective and caused harm. This is a materially different (and easier for plaintiffs) standard than ordinary negligence law, and it is a key reason US product liability exposure is larger than in many other jurisdictions.

The EU runs a comparable strict liability framework under the Product Liability Directive, recently updated (the new directive entered into force in 2024, with member states transposing it into national law through 2026), which extends liability explicitly to software and AI-enabled products, an important shift for manufacturers embedding connected components.

Civil suits can proceed as class actions, aggregating thousands of harmed customers into one case, which is what happened with Takata and with GM's ignition switch defect (2014, at least 124 deaths linked, over $2.5 billion in combined recall, settlement, and penalty costs across criminal and civil resolutions).

Where the three systems meet, and where they clash

The regulator, the reporting clock, and the courts do not always move together, and that gap is where most of the business risk sits.

SystemWho runs itMain sanctionTiming
CPSC / NHTSA / FDAFederal agencyMandatory recall, finesCan act fast once notified
Mandatory reportingStatute (CPSA, TREAD Act)Separate penalty for late disclosureStarts the moment internal knowledge exists
Tort liabilityState courtsDamages, punitive damagesCan run for years after a recall closes

A company can complete a recall, satisfy the regulator, and still face a decade of individual and class action lawsuits. Takata's bankruptcy did not end Honda's or other automakers' litigation exposure; it shifted the financial burden onto the automakers that installed the defective part, because product liability can attach anywhere in the supply chain, not just at the original manufacturer.

This is the central trade-off framed in the hook: launching faster (or delaying disclosure of a known issue) reduces short-term cost and preserves market position, but it compounds downstream exposure across three independent penalty systems that don't forgive each other's mistakes.

🎬 [VIDEO: "How the Takata Airbag Recall Became the Largest in US History" - youtube.com - search for NHTSA or major news retrospectives explaining the defect mechanism and recall scale]

Knowledge check

1. Why does the Takata airbag case illustrate the need for manufacturers to understand multiple overlapping safety systems rather than just one?

2. A company manufactures a component that is installed in motor vehicles. Which regulator would have primary jurisdiction over a safety defect in that component?

3. What is the key distinction between OSHA's role and the roles of CPSC, NHTSA, and FDA in the product safety landscape?

MULTIPLE CHOICE

4. Select ALL correct answers about how US federal product safety regulation is structured.

Select all the correct answers.

MULTIPLE CHOICE

5. Select ALL correct answers about the significance of the Takata airbag case for manufacturing executives.

Select all the correct answers.

What compliance actually looks like on the ground

For a manufacturing professional, this isn't abstract law, it shows up as operating procedure:

  • Field data monitoring: warranty claims, dealer complaints, and customer service logs must be systematically reviewed for defect signals, because "we didn't know" is a weak defense once a pattern existed in the data.
  • Document retention: internal test results and engineering emails become discoverable evidence in litigation; Takata's own internal test data was central to prosecutors' case.
  • Recall execution: once a recall is announced, tracking remedy completion rates matters. Airbag recalls have historically had completion rates well below 100% for years after issuance, since owners don't always respond to notices, a persistent regulatory concern documented in NHTSA's recall completion data.
  • Supply chain contracts: indemnification clauses determine who ultimately pays when a defective component (like an inflator) causes harm to an end product (a car); this is a live negotiation point in every tier-1 supplier contract.

Key Takeaways

  • Manufacturing safety runs on three separate, overlapping systems: a regulator (CPSC, NHTSA, FDA, or EU equivalents like GPSR/Safety Gate), a mandatory reporting clock (24 hours under CPSA Section 15(b)), and civil tort liability decided in courts under strict liability standards.
  • Delayed internal disclosure of a known defect is a separate legal violation from the defect itself, as Takata's $1 billion criminal penalty demonstrated.
  • Passing a recall does not end liability exposure; class action litigation can run for years after regulatory closure, and costs can shift down the supply chain to whoever installed the defective component.
  • The EU's updated Product Liability Directive (in force 2024, transposing through 2026) now explicitly extends strict liability to software and AI-enabled product features, a critical point for manufacturers adding connected or smart components.
  • Practical compliance means continuous field-data monitoring, disciplined document retention, and clear indemnification terms in supplier contracts, not just a one-time regulatory filing.

Frequently asked questions

Who pays when a defective product injures someone?

Liability can attach anywhere in the supply chain, not only to the company that made the faulty part. When Takata's airbag inflators ruptured, Takata went bankrupt and the financial burden shifted to automakers like Honda, which had installed the inflators and paid billions in recall costs, settlements and fines. Supplier indemnification clauses usually decide who absorbs the final bill.

How fast must a manufacturer report a product defect to the CPSC?

Within 24 hours. Section 15(b) of the Consumer Product Safety Act requires manufacturers, importers, distributors and retailers to notify the CPSC within 24 hours of obtaining information reasonably supporting the conclusion that a product has a defect creating a substantial risk of injury or violates a safety standard. For vehicles, the TREAD Act gives manufacturers 5 business days to notify NHTSA once a safety defect is determined.

What is the difference between a design defect, a manufacturing defect and failure to warn?

A design defect means the product is unreasonably dangerous as conceived, even when built perfectly, as with Takata's ammonium nitrate propellant. A manufacturing defect means a specific unit deviated from the intended design, such as a contaminated batch. Failure to warn means the product carries an inherent risk in some use case that the manufacturer did not adequately disclose. These are the three theories that dominate US product liability suits.

Does completing a recall end a manufacturer's legal exposure?

No. A company can finish a recall, satisfy the regulator and still face years of individual and class action lawsuits, because civil courts run on a separate track from safety agencies. Takata's bankruptcy did not close out litigation for the automakers that had installed its inflators. Regulatory closure and tort liability are independent outcomes.

Why is US product liability exposure larger than in other jurisdictions?

Mostly because of strict liability: in most US states a plaintiff only has to show the product was defective and caused harm, not that the manufacturer was negligent. Combined with class actions that aggregate thousands of claimants and punitive damages, this produces far bigger settlements than negligence-based systems. The EU applies a comparable strict liability framework through its Product Liability Directive, but without the same litigation machinery.

Does the new EU Product Liability Directive cover software and AI features?

Yes. The revised directive, in force since 2024 and being transposed into national law through 2026, extends strict liability explicitly to software and AI-enabled products. For manufacturers adding connected or smart components to physical goods, a faulty algorithm or firmware update can now trigger the same liability as a faulty mechanical part.

Related articles

Recent articles from the blog that build on this lesson.