Skip to content 
Artificial intelligence is becoming embedded in everyday business operations, from customer service chatbots and automated decision-making to document drafting and data analysis. As AI adoption increases, one question has become increasingly important: who is legally responsible when AI causes harm?
This issue has recently been brought into sharp focus following the publication of the UK Jurisdiction Taskforce’s Legal Statement on Liability for AI Harms under the private law of England and Wales. The statement confirms that existing English law is capable of dealing with many AI-related disputes without the need for entirely new legislation.
AI is not legally responsible – people are
A common misconception is that AI itself could be liable for mistakes or harmful outcomes.
Under English law, AI has no legal personality. This means that responsibility will generally rest with the individuals or organisations that develop, deploy or rely upon AI systems.
Depending on the circumstances, liability may arise under established areas of law including:
- negligence;
- breach of contract;
- misrepresentation;
- product liability;
- defamation; and
- data protection legislation.
Rather than creating entirely new legal principles, the courts are expected to apply existing legal doctrines to AI-related disputes.
Examples of where liability may arise
Businesses should be aware that AI-generated errors can create significant legal exposure.
Examples include:
- an AI chatbot providing incorrect contractual or financial advice to customers;
- automated recruitment software discriminating against job applicants;
- AI systems generating defamatory content;
- AI making inaccurate representations about products or services;
- AI tools causing financial loss through inaccurate calculations or recommendations.
In each case, liability is likely to depend upon whether the organisation acted reasonably in selecting, implementing and supervising the technology.
Human oversight remains essential
AI should not be viewed as removing human responsibility. Businesses should ensure appropriate governance arrangements are in place, including:
- staff training;
- clear AI usage policies;
- regular auditing of AI outputs;
- maintaining human review for high-risk decisions; and
- documenting how AI systems are used within the organisation.
The greater the potential impact of an AI decision, the stronger the expectation that meaningful human oversight will be maintained.
What should businesses do now?
Although AI-specific legislation continues to evolve, organisations should not wait for new laws before reviewing their legal risks.
Practical steps include:
- reviewing contracts relating to AI suppliers;
- implementing internal AI governance policies;
- ensuring compliance with data protection obligations;
- identifying where human review is required; and
- seeking legal advice before deploying AI in customer-facing or high-risk environments.
