Introduction

You are at the wheel of a brand-new car hurtling down Germany’s autobahn on cruise control at 120 kilometers per hour. You are running late for a client meeting and talking on the phone with a colleague. Another client is asking for a Teams meeting, as you drive. Your attention to driving is, at this point, minimal. A car ahead of you suddenly pulls into your lane, to avoid debris on the road. Thankfully, although your focus is elsewhere, your new car remains avidly attentive to your trip. The car warns you and a potential catastrophe is averted. You think nothing more of it as you try to make up time on your way to see your client. Yet, you are riding in a complex automotive system created by sensors, computer chips, cameras, cruise control, brakes and steering components. These systems monitor the location and speed of adjacent vehicles. The systems warn when you stray out of your lane or approach another vehicle ahead of you with too much speed.

Digital components increasingly determine the safety and functionality of modern products.

European law on product liability stood unchanged for forty years. Since 1985, product liability litigation in Europe followed the European Union’s (“EU”) Council Directive 85/374/EEC. Why did the EU Parliament choose to overhaul product liability law now? Here’s your answer: an automobile mechanic in 1985 changed spark plugs and cleaned carburetors. There was not a single computer in an automobile. Indeed, personal computers were just coming on the market. They were clunky and slow. The European Parliament recognized that we now live in a different era. As a result, it significantly updated and revised its product liability regime. Directive (EU) 2024/2853, was issued on 8 December 2024 and must be transposed by Member States by 9 December 2026. The Directive can be found at [eur-lex.europa.eu]. Every European country is now obligated to conform its existing laws to the new directive by the December deadline. If a country falters in conforming its existing laws, fines may ultimately result.

No matter where you are located in the world, if your clients include manufacturers, distributors, importers, insurers, tech startups, fulfillment services, or individuals who were injured by a product, you need to understand the new changes. They are substantive and require careful attention, so that you will be prepared to correctly advise your clients on what to expect and how to avoid pitfalls as they chart a course ahead.  

The new Directive significantly expands the reach of product liability law to address software, artificial intelligence (AI), digital products, circular economy business models, and complex global supply chains. The reform also eases evidentiary burdens and broadens recoverable damage for individuals who claim that they have been injured by a product. There is much to discuss. Let’s get started.

How the New Directive is Different

The 1985 Directive was designed for tangible consumer products. This included every type of consumer product, from blenders to truck tires. But the 1985 Directive was issued long before the technology revolution created a dot.com economy powered by smartphones, laptops, autonomous vehicles, drones and robots. Most of the risks that bedevil current manufacturers were undreamed of in the 1980s. Imagine taking the auto mechanic mentioned above, who was trained to repair 1985 cars with spark plugs and now asking that mechanic to repair a modern Tesla. That would probably not end well for the mechanic or the Tesla owner. The proliferation of cloud-connected services, cybersecurity risks, software as a service, AI powered products and post-sale software updates, caused EU regulators to reexamine: (a) what constitutes a “product;” (b) who should bear responsibility when software and related components cause harm; and (c) how liability should be determined.

The EU legislature also examined the challenges consumers may face in obtaining and understanding technical evidence and proving causation in technologically complex cases. The new Directive attempts to “modernize” product liability law both substantively and procedurally. As the example of our spark plug-wielding mechanic shows, new products and concepts necessarily call for changes in the law of product liability. But questions abound. Do the EU’s changes tilt the playing field too much in the favor of claimants? Or is the law a prescient and necessary overhaul of antiquated concepts? Are centuries-old judicial concepts on proving a case in court being casually discarded? Is Europe’s traditional abhorrence of American style discovery evaporating? If so, will the new Directive have far-reaching and unanticipated effects? Will it stifle creativity by applying onerous burdens to defendants who must learn to litigate under the new laws? Will the global reach of the directive substantially slow and hamper the introduction of innovative products produced both inside and outside Europe? Does Silicon Valley need to fear this law as much as a tractor manufacturer in Bavaria? It may take another 40 years before the implications of the law are fully understood and, doubtless, the law will be changed by then, but in the interval, we will examine the most salient, controversial, and intriguing changes.   

What is a Product?

As mentioned above, the 1985 Directive applied to tangible goods, ranging from toasters to pharmaceutical medicines. The new Directive expressly and purposefully expands the definition of a product to include software components, cloud software, AI systems, and digital manufacturing files used for 3D printing. Both stand-alone software and software integrated into physical products are included. The Directive does carve out and exclude certain non-commercial, free and open-source AI software. The Directive is also relevant for businesses involved in substantial product modifications and software updates, because it recognizes that such updates may unwittingly generate liability, even if the original product was not defective when it was sold.  

A Broader Range of Potentially Liable Parties

The 1985 framework focused principally on the actions of manufacturers and importers into the EU.

Now, companies that previously viewed themselves as intermediaries may find themselves subject to greater litigation risk. The new Directive broadens the range of businesses who may be exposed to liability. Modern supply chains draw components from around the globe, and a consumer may not be able to determine where a particular component originated. Think of a computer chip embedded in a refrigerator.

When a manufacturer cannot be identified or is located inside the EU, the Directive permits the extension of liability to authorized representatives, as well as online platforms, fulfillment services and distributors. As an example, Amazon’s “Fulfillment by Amazon” service aids manufacturers by storing inventory, picking, and packing a requested product, shipping it to a customer and processing returns. The Directive acknowledges the new reality that a manufacturer in Vietnam or Brazil may use a fulfillment service to make its product available through online channels in Europe.

Further, the Directive specifically includes defects caused by cloud computing. This means that companies that operate cloud services, such as Amazon, Microsoft, Google, IONOS, and STAKIT, should assess and plan for increased liability risk.   

Expanded Categories of Recoverable Damage

The new Directive broadens the scope of compensable harm. Like its predecessor, the Directive covers death and personal injury. However, it now expressly recognizes medically diagnosed psychological harm as the basis for recoverable damage.

In addition, the Directive introduces protection for data destruction. In an era where data loss may have significant economic and personal consequences, this represents a notable departure from the traditional focus on physical property damage.

The reform also removes the previous €500 threshold for property damage claims under the 1985 Directive, thereby lowering barriers to recovery for claimants. Finally, individual EU member countries will no longer be permitted to cap or limit liability for personal injury or death. And pain and suffering damages will be allowed, if permitted by individual members. The Directive also extends compensation not only to product users who allege direct damage but also indirect claimants who allege injury based on the direct user’s injury. Such indirect damages may include wrongful death claims and loss of support from an injured user.

New Rules on Product Defectiveness

Although the fundamental test remains whether a product provides the level of safety that the public is entitled to expect, the new Directive identifies additional factors relevant to the assessment of defectiveness.

Courts may now consider issues such as cybersecurity vulnerabilities, software updates, machine learning capabilities, foreseeable misuse, and the product’s ability to continue operating safely throughout its lifecycle. Some products, such as medicines and life-sustaining medical devices entail a high risk of causing damage and therefore, according to the drafters, create particularly high safety expectations. Specifically, the level of safety the public is “entitled to expect” is correspondingly higher. But science and technology never advance in a perfectly linear manner. Rather, they advance in fits and starts. There are many reasons why someone may incur an injury from a product, such as a new medicine or medical implant. Will heightened design scrutiny, coupled with a presumption of defect, create unintended disincentives to innovation? Will manufacturers choose to leave an earlier product on the market rather than risk liability for bringing a new product to the market? Is it appropriate to ask if heightened scrutiny always advances the public’s interest?

The Directive also indicates that liability cannot be avoided merely by listing all the conceivable side effects of a product. In other words, warnings alone cannot make a defective product safe. Instead, the safety expectations of the public  controls. But how does a court evaluate the “expectations of the public at large?” We know that a court will consider warnings, foreseeable use (and misuse), state of the art, compliance with product regulations and recalls. Yet, at the end of the day such an evaluation seems not fully objective but rather at least partially subjective.

The Directive creates another conceptual shift. Product safety is no longer assessed solely at the moment the product is first placed on the market. Ongoing updates, upgrades, and digital maintenance may influence liability exposure long after initial distribution.

Enhanced Disclosure Obligations and Presumptions

The Directive is plainly intended to ensure that consumers receive compensation for defective products and one of the most claimant-friendly aspects of the reform concerns evidence. Europe has traditionally eschewed the type of extensive, time-consuming and expensive pre-trial discovery that characterizes the American tort system. In Europe, claimants must present the evidence on which they intend to rely and requests for document production must be specific and surgically targeted. “Fishing expeditions,” in which a claimant requests broad categories of documents, are usually not allowed. The Directive upends these traditional discovery guardrails.

The drafters perceived that claimants, without the benefit of extensive pre-trial discovery, were at times handicapped in obtaining technical documentation that might help them establish defectiveness and causation. The new Directive therefore allows courts to order disclosure of relevant evidence held by defendants in appropriate circumstances. Such circumstances may include the scientific or technical complexity of a product. Additionally, a defendant may now have to disclose not only design specifications but also safety risk assessments and reported potential product defects.

In addition to the discovery upheaval, counsel who regularly represent product defendants may be surprised and disappointed to learn that the Directive also establishes rebuttable presumptions relating to defectiveness and causation in certain situations, particularly where claimants face “excessive” evidentiary difficulties due to technical or scientific complexity. Such presumptions might apply to pharmaceutical drugs and complex, computer-run machinery and systems, such as wind turbines. But don’t such presumptions erase decades of jurisprudence, in which the claimant had the burden of establishing a defect, as well as causation? And how does a defendant go about rebutting a presumption that a product is defective, simply because it is complex? Are not most new and advanced products based on complex science and engineering principles? Will courts now throw up their hands and say a product is too complex for a common person to understand, so if the product injures someone, it must be defective? Are we about to go down a slippery slope, in which well-designed and manufactured products are nonetheless, automatically presumed defective?     

For litigators, these procedural reforms may prove just as significant as the substantive expansion of liability.

Implications for AI and Emerging Technologies

Although the Directive does not include a standalone AI liability section, it clearly brings AI-enabled products within the framework of strict product liability.

The Directive  was expressly motivated by technological developments, including artificial intelligence, and aims to ensure that consumers receive comparable protection regardless of whether harm arises from traditional mechanical products or sophisticated digital systems.

Organizations developing AI-enabled products should therefore review governance, testing, documentation, cybersecurity controls, update procedures, and record-keeping practices to mitigate future litigation risks.

Conclusion

EU Directive 2024/2853 represents the most significant reform of European product liability law in four decades. While preserving the strict liability foundation established by Directive 85/374/EEC, it substantially expands the concepts of product, damage, and potentially liable parties. At the same time, it strengthens claimants’ procedural tools through disclosure obligations and evidentiary presumptions.

For attorneys around the globe, the transition period before the December 2026 implementation deadline provides a critical window to advise clients on compliance, risk allocation, contract drafting, indemnity agreements, product governance, and litigation preparedness. As software, AI, and connected technologies become increasingly central to modern commerce, the new Directive ensures that EU product liability law is no longer anchored solely in the physical products of the twentieth century.

written by William E. Vita, partner in the New York office of Shook, Hardy & Bacon L.L.P.. He has deep experience in product liability law and regularly represents clients located in both Europe and the United States