The UK Jurisdiction Taskforce (UKJT) published its Legal Statement on Liability for AI Harms under the Private Law of England and Wales on 7 July 2026. The statement is final, produced following a public consultation that ran from 14 January to 13 February 2026. The UKJT is an industry-led body operating under LawtechUK, tasked with clarifying how English law applies to digital technologies. The statement does not create binding precedent but is an authoritative analysis that courts and arbitrators are likely to consider when resolving AI liability disputes under English private law.
The statement analyses liability under five doctrines. In contract, AI involvement creates no special difficulty: contractual terms allocate liability for AI-caused harm, and rules on implied terms, fitness for purpose under the Supply of Goods and Services Act 1982, and limitation of liability apply without modification. In negligence, the Caparo three-stage test applies: duty of care, breach of the applicable standard, and causation of foreseeable harm. Vicarious liability attaches to employers where a human employee caused harm through wrongful use of AI in the course of employment. Under the Consumer Protection Act 1987, software and AI models are not currently 'products' within the Act's definition, leaving a gap the UKJT recommends Parliament address.
The statement carries specific implications for three categories of market participant. Foundation model developers are, in most circumstances, unlikely to face negligence liability for harm caused by unforeseeable uses of their general-purpose models by third parties. Organisations that present an AI chatbot as communicating on their behalf, or that adopt AI-generated statements as their own, bear liability under agency and misrepresentation principles for those AI outputs. Professionals, including lawyers, financial advisers, medical practitioners, and engineers, may face professional negligence claims for negligent use of AI, and also for failing to use AI where a competent member of their profession would have done so.
The statement leaves several questions open. It does not address criminal liability or regulatory liability under the Financial Services and Markets Act 2000, the UK GDPR, or the Online Safety Act 2023. The analysis applies to autonomous AI systems and does not resolve liability for human-AI collaborative outputs where human and machine contributions are difficult to separate. The statement identifies the software product liability gap under the Consumer Protection Act 1987 as a significant issue and recommends legislative reform.
Licentium advises businesses deploying AI on liability exposure under English private law, contractual risk allocation in AI procurement and supply chain agreements, and regulatory obligations under the FCA and ICO. If you are reviewing AI contracts, operational governance, or professional indemnity exposure in light of the UKJT statement, we can assist. Work we undertake includes AI deployment risk reviews, contract drafting and negotiation for AI procurement, professional liability analysis, and advice on AI governance under sector-specific regulatory regimes.