Who is liable when AI goes wrong? A new series examining AI liability and risk from a global perspective

Artificial intelligence is increasingly embedded in business operations, decision-making and customer-facing services. As adoption accelerates, organisations are understandably focused on the opportunities AI presents. But alongside those opportunities comes a growing need to understand and manage the risks.

In this series, we explore how legal systems and regulators are responding to the challenges presented by AI, and what those developments mean in practice for businesses and insurers. Looking at developments in the UK, Europe, Middle East and the United States, we will examine different approaches to liability, risk allocation and governance, before considering the practical measures organisations can take to manage AI-related risks before harm occurs.

Article 1: The UK Jurisdictional Taskforce consultation and European developments

The first article will examine the UK Jurisdiction Taskforce's recent consultation on liability for AI harms under the private law of England and Wales and compare it with the EU's emerging legislative framework. We will explore the UKJT's conclusion that existing principles of contract, negligence and product liability are generally capable of addressing AI-related harms without a bespoke AI liability regime, and contrast this with the EU's more interventionist approach through the AI Act and revised Product Liability Directive. The article will also consider the practical implications of these developments for businesses and insurers, particularly in relation to risk allocation, governance and cross-border exposure.

Article 2: AI liability and emerging litigation 

The second article will consider recent AI-related litigation in various jurisdictions, including claims arising from allegedly harmful chatbot outputs and AI-generated misinformation. Using these cases as practical examples, we will examine whether AI is creating genuinely novel legal risks or whether courts are largely applying established negligence, product liability and misrepresentation principles to new technologies. We will also consider the extent to which the themes emerging from the UKJT consultation and the revised EU Product Liability Directive provide a useful framework for understanding these disputes, before assessing the resulting implications for businesses and insurers.

Article 3: AI liability in the Middle East 

As governments across the Middle East pursue ambitious AI agendas, particularly in the UAE and Saudi Arabia, businesses, regulators, and courts are increasingly grappling with a fundamental question: who is liable when AI goes wrong? This article will explore the emerging litigation risks associated with AI systems, from algorithmic errors and autonomous decision-making to data misuse and AI-generated harm; examine how existing legal concepts such as negligence, product liability and regulatory compliance may evolve to address these challenges; and consider how the legal framework here in the Middle East differs from other jurisdictions. Against a backdrop of fast-moving regional AI initiatives and developing regulatory frameworks in the Middle East, it will consider how liability for AI-related harm is likely to be allocated among developers, deployers and users in the years ahead.

Article 4: AI liability in the United States 

This fourth article examines the evolving U.S. legal framework for AI, where regulation is driven by a combination of federal agency enforcement, executive policy, existing laws, and a growing patchwork of state legislation rather than a single federal AI statute. The article will address litigation trends, including consumer and employment-related class actions alleging algorithmic bias, privacy violations, and deceptive AI practices. The article also analyses the rapidly developing body of copyright and training data litigation against AI developers over the use of copyrighted works to train foundation models. Together, these developments are shaping the contours of AI governance, exposure, and liability in the U.S.  

Article 5: AI: contractual, risk and governance considerations 

The final article will focus on the practical measures organisations can take to manage AI-related risk before harm occurs. This article explores identification of risk and risk management at different stages of the AI lifecycle and how to address it from a contractual and compliance perspective, without slowing business as usual. As organisations continue to rely on the advantages use of AI can provide, our experts explore how organisations should approach AI governance frameworks, legal compliance, legal risk mitigation and internal controls.

Conclusion

Taken together, the series will explore how different legal and regulatory frameworks are approaching AI-related harms, how liability may be allocated when things go wrong, and what those developments mean for businesses and insurers. Follow along as we examine the evolving approaches to AI liability in the UK, Europe, Middle East and the United States, before considering the practical governance, oversight and risk management measures that may help organisations navigate these issues.