Product Liability

You Waited for Transposition. Now What?

The revised Product Liability Directive applies to products placed on the market after 8 December 2026. The most useful test now is whether you can reconstruct one of them.

EU-themed Readiness Directive banner showing product records, regulatory evidence and a product reconstruction concept ahead of the revised PLD.
EU-themed Readiness Directive banner showing product records, regulatory evidence and a product reconstruction concept ahead of the revised PLD.

The revised Product Liability Directive applies to products placed on the market after 8 December 2026. The most useful thing to do before then is to find out whether you can reconstruct one of them.

For almost two years, waiting was a defensible position. Directive (EU) 2024/2853 had been adopted, but the national laws that would carry it into courtrooms were still being drafted. Procedure, where much of the practical difference sits, remained unclear. Many legal teams reasonably chose to see the national text before asking the business to redesign anything.

That position now has a fixed end. Member States must bring their implementing measures into force by 9 December 2026 (Article 22). Following the May 2026 corrigendum, the revised regime applies to products placed on the market or put into service after 8 December 2026. Products placed earlier remain under Directive 85/374/EEC (Article 21).

Nothing dramatic happens in December. No wave of judgments arrives, and the first claims under the new rules are likely to be some way off. What does happen is quieter, and it matters more to anyone who waited. The first products that will be judged under the revised Directive are being documented now, by people who are not thinking about litigation. Their firmware is being frozen, their supplier concessions signed and their risk assessments closed.

So the remaining weeks are best spent on a practical test. Take one product that will enter the new regime and find out whether Legal, Quality, Engineering and Product Safety can reconstruct three things: what it was, what the organisation knew about it, and who decided what.

The reconstruction has to be in a form someone outside the business could follow.

Most organisations have the records. Far fewer have tried to rebuild a product from them, and the Directive makes that gap matter.

What Article 9 asks of a defendant

Article 9 allows a claimant to ask a national court to order the defendant to disclose relevant evidence at its disposal once the claimant has presented facts and evidence sufficient to support the plausibility of the claim. Disclosure is limited to what is necessary and proportionate. The court must weigh the legitimate interests of all parties, including third parties, particularly where confidential information and trade secrets are involved.

Article 9(2) also gives defendants a reciprocal route to evidence held by claimants.

The mechanism is narrower than US-style discovery. Recital 43 leaves questions including pre-trial procedure, the specificity of requests, evidence held by third parties and sanctions to national law.

Two provisions deserve particular attention.

Article 9(6) empowers national courts, in accordance with national law and where proportionate in cost and effort, to require disclosed evidence to be presented “in an easily accessible and easily understandable manner.”

Recital 42 adds that evidence a defendant may have to disclose can include documents created ex novo by compiling or classifying evidence it already holds.

My reading is that these provisions change the practical question an organisation should ask.

An archive can answer:

Do we have the documents?

Only people who understand the product, its records and its decisions can answer:

Could we compile them into an account of this product that a court could follow?

Most organisations have never asked them to try.

Run one reconstruction drill

Pick one product the organisation expects to place on the market after 8 December 2026.

A mature product with real field history is a better choice than a showcase launch because it carries the accumulated complexity that a claim would expose.

Give Legal, Quality, Engineering and Product Safety a fixed window of a few days and ask them to establish:

  • the exact product, variant and configuration supplied;
  • the software or firmware state, and how that is known;
  • the design assumptions and risk assessments that applied;
  • the supplier deviations and concessions accepted;
  • the instructions and warnings in force;
  • what was known from the field, and when;
  • the changes made after release, and who approved them.

Then ask whether those findings can be assembled into one coherent account.

Finding fifteen repositories is not the same as reconstructing one product.

The technical file, complaint database, software repository, service platform, supplier portal and email archive may each be accurate and still fail to explain the product together. They use different identifiers, describe different versions and were never designed to be read side by side.

The friction tends to sit in predictable places.

In fire and security systems, for example, the behaviour of an installed product can depend on configuration set on site by an installer or integrator. That configuration may sit in commissioning records the manufacturer does not hold.

Elsewhere the obstacle is an archived system nobody can query, a decision recorded only in email, or an engineer who carried the reasoning in their head and has since left.

Record four things:

  • how long the reconstruction took;
  • what could not be found;
  • what depended on a single person's memory;
  • what could not be explained to someone outside the function that created it.

If Legal can establish the configuration only by calling six functions and tracking down a retired engineer, the weakness already exists.

A claim would simply reveal it.

Where Article 10 makes weak evidence expensive

Article 10 keeps the burden of proving defectiveness, damage and causation on the claimant. It then establishes rebuttable presumptions for defined circumstances.

Defectiveness is presumed where:

  • the defendant fails to disclose relevant evidence under Article 9(1);
  • the claimant demonstrates that the product does not comply with mandatory product-safety requirements intended to protect against the risk of the damage suffered;
  • the damage was caused by an obvious malfunction during reasonably foreseeable use or under ordinary circumstances.

Recital 46 makes clear that the safety-requirements route can include a product that lacks legally required capability to log information about its operation.

Once defectiveness is established, causation is presumed where the damage is of a kind typically consistent with that defect.

Article 10(4) deals with a different situation. Even after disclosure, and taking account of all the circumstances, a court must presume defectiveness, causation or both only where the claimant faces excessive difficulties in proving them, particularly because of technical or scientific complexity, and also demonstrates that defectiveness or causation is likely.

Every presumption can be rebutted.

Most claims will therefore not turn on a presumption, and the claimant's burden of proof remains fundamental.

The operational point is narrower.

An organisation that cannot find what it holds is badly positioned to demonstrate compliance with a disclosure order. One that cannot reconstruct the relevant product state may also struggle to rebut a presumption once the conditions for one have been met.

The reconstruction drill tests both.

Five things to do before December

None of these requires a transformation programme.

Each should produce something measurable and change a decision.

This article provides independent analysis and is not legal advice. Regulatory status and dates should be verified against current official sources.