First sight of landmark proposals to reshape product liability
On 8 October 2026, the Law Commission of England and Wales published its consultation on reforming the product liability regime under the Consumer Protection Act 1987 (CPA) (see our blogs here and here for background). The consultation closes on 14 January 2027, with final recommendations expected in late 2027 (consultation paper here; project documents here).
The publication is against the backdrop of EU Member States transposing the revised Product Liability Directive (EU) 2024/2853 (PLD2) into national laws before the 9 December 2026 deadline (see some of our analysis of PLD2 here and here). The Law Commission’s provisional proposals reflect many aspects of PLD2, but with important differences.
Key proposed liability reforms
The Law Commission has made provisional proposals for change, and in some cases asked open questions, across most of the core elements of the CPA, which we summarise below.
Expanded definition of product
The Law Commission provisionally proposes extending the definition of product to software and AI systems, including where embedded in hardware or accessed through a subscription model. Taking the definition beyond the approach of the EU, it would also include ‘the design components and features of social media platforms’. The definition would also be future-proofed by being extended ‘to any other thing (digital or otherwise) that is designed or developed to perform a specific function for users, irrespective of how it is supplied or made available’.
New categories of defendants
Under the provisional proposals, the definition of producer used in the CPA would be extended and there would be three additional categories of potential defendants that could be sued in certain circumstances. These are: (i) fulfilment service providers; (ii) ‘core participants’ (a broad category covering entities that have ‘played a significant role in bringing a product to the UK market’, including online marketplaces); and (iii) UK authorised representatives (persons authorised to act on another person’s behalf in relation to the product, including for regulatory and market-access functions). Fulfilment service providers and core participants would only be liable where no UK producer, own-brander, importer or authorised representative can be identified. The aim of having these different tiers of defendants is to increase the likelihood that the claimant will be able to identify a defendant within the jurisdiction.
Test for ‘defect’
The essential test for defect – that the product does not present the level of safety which the public is entitled to expect, in all the circumstances - would not change. However, the provisional proposals seek to provide greater guidance to courts on assessing defect by reference to seven non-binding principles. These principles include assessing a product’s risks and benefits, which is a standard exercise conducted by businesses in certain tightly regulated sectors (eg pharmaceuticals). The Law Commission acknowledges this in its consultation, recognising that claimants may find it difficult to rely on the principle as a ground for finding defect where a product has passed a formal risk-benefit assessment, particularly where that is a pre-condition of marketing. However, the consultation notes that such prior assessments may have been based on incomplete or inaccurate information, and that the court would therefore not be bound to reach the same conclusion.
There would also be additional statutory circumstances that the court would be obliged to consider, including post-supply learning or new features and compliance with relevant cybersecurity requirements, standards and codes of practice.
Evidentiary presumptions
A core feature of the CPA (and both the first Product Liability Directive that it implemented, and PLD2) is that a claimant does not need to prove fault on the part of the producer, only that a product was defective and that the defect caused them damage falling within the scope of the Act. Taking this further, and following aspects of PLD2, the Law Commission provisionally proposes a rebuttable presumption that the product is defective in certain circumstances. This would apply where the claimant satisfies the court that it is arguable that: (i) the product is defective; (ii) the facts relied on are not fanciful; (iii) the claimant faces exceptional evidential difficulties; and (iv) the risk of harm is not so widely known as to be ‘socially accepted’. If these criteria were met, the defendant would then choose either to assume the burden of rebutting the presumption of defect or to accept the alleged defect. Where accepted (and subject to proof of causation and no available defences), the court would apportion the risk between the parties with a starting point of a 50% split of the recoverable damages for each party. On the other hand, if the defendant chose to attempt to rebut the presumption of defect and failed, the defect would be treated as established and damages would assessed as they would be if defect had been proved. The claimant would still need to prove damage and causation.
The Law Commission also poses an open question on whether to introduce a rebuttable presumption of a fact relevant to defect and/or causation when a claimant can show that there is an asymmetry of information and that it would be fair for the defendant to bear the burden of rebutting the presumption. It is unclear whether the Law Commission envisages that this and the ‘exceptional difficulties’ presumption could be combined; if so, then this could make it practically difficult for defendants in relevant cases.
Defences
The CPA’s existing defences would remain under the provisional proposals, but they would be adapted, mainly to reflect digital products. This includes narrowing the defence based on showing that the product was not defective at the time of supply (factory gate principle) where a defect arises from matters within the defendant’s control after the supply date (for example, software updates or related services). The Law Commission also proposes retaining the development risks defence, which protects a producer where the state of scientific or technical knowledge at the time of supply was not such that a producer of similar products might be expected to have discovered the defect. Recognising that other aspects of its provisional proposals seek to extend the scope of liability for businesses, the Law Commission describes the defence as ‘necessary mechanism to balance, fairly, business confidence and innovation against consumer and third-party protection’.
Recoverable damage
The CPA allows compensation for death, personal injury and damage to private property, with property claims limited to losses above £275. The Law Commission proposes removing this threshold and extending recovery to property used for both private and professional purposes. In addition, again similar to PLD2, the destruction or corruption of data would be recoverable. The Law Commission has also provisionally proposed that losses incurred when taking necessary measures to avert a risk of personal injury should constitute compensable damage.
Most significantly, and in line with PLD2, the Law Commission recommends that damages for recognised psychiatric illness should be recoverable, in the absence of physical injury (which would currently only be the case in very rare instances). This opens the door to claims alleging addiction or other psychological or psychiatric conditions allegedly caused by exposure to a (physical or digital) product. Recoverable damage would include psychiatric harm suffered by indirect victims (ie victims who have close ties of love and affection with a direct victim) in some cases.
Longstop
Currently, a claimant must issue their claim form within ten years of the date of supply of the product in order for it to be permitted to proceed. The Law Commission’s proposals maintain this ten-year ‘longstop’ in most cases but provide for important exceptions. These include resetting the longstop where a product is substantially modified after supply, for example through a software update. In addition, as under PLD2, the longstop would also be extended to 25 years for latent personal injury, where it was not possible for the claimant to start their claim within the ordinary ten year period. The Law Commission then goes further, by proposing that courts should be able to substitute a defendant after the longstop has expired, for example where a claim has been issued against the wrong company in a corporate group. The CJEU had previously held that this was not permissible under the current regime.
Comment: Striking the balance between consumer protection and innovation
These reforms are comprehensive and represent a thoughtful attempt to balance the objectives of consumer protection and innovation, as well as the challenges posed by emerging technologies. However, the underlying premise of many elements of the reforms appears to be that the current regime is tilted significantly in favour of defendants. There is limited, if any, empirical evidence that this is the case; examples of claimants failing to bring successful claims do not equate with a flawed system. There is a real risk that these proposals, if they became law, would tilt the balance too far against defendants.
It is also unclear that all these changes are necessary. The common law may also have allowed for incremental legal development to achieve the CPA’s objectives without the broad reforms envisaged by the Law Commission. This was the sentiment of the UK Jurisdiction Taskforce, an established group of experts from the judiciary and legal profession, in its legal statement on liability for AI harms (see our blog here and article here).
Looking ahead
After the consultation closes on 14 January 2027, the Law Commission expects to publish final recommendations in late 2027, potentially accompanied by draft legislation. After that, we will likely see a Government response to the proposals and clearer legislative proposals. In the past, UK governments have typically followed Law Commission recommendations (at least in terms of their broad thrust) in around two-thirds of cases. It is therefore important for stakeholders to engage both with this process and with parallel initiatives such as the Law Commission’s work on consumer class actions (including possible opt-out consumer mass claims).
