ABSTRACT
Artificial intelligence (AI) is now deeply integrated into our daily lives. More and more people are relying on tools such as ChatGPT and other legal AIs to help them understand the law, navigate contracts and seek legal advice. While these AIs can be very helpful, they are not always completely accurate, often making up information that sounds correct but is completely false called an ''AI hallucination''. However, if an individual takes an action based on these fabricated yet realistic statements and is harmed by it, who is liable for such hallucination? Is it the company providing such AI liable? The lawyer using it? Or even the individual themselves.
This paper discusses the concept of AI hallucinations as a source of defective legal advice and the legal principles that could be used to govern liability. The paper argues that although AI is a valuable tool that can enhance the efficiency of legal practice it is not a substitute for professional legal opinion. The paper will review the legal principles, case law and ethical frameworks relevant to the issue and will conclude with recommendations on the most appropriate way of governing liability to ensure that legal practice remains ethical and that the public is protected from the risks associated with the use of such technologies.
KEYWORDS
Artificial Intelligence, AI Hallucination, Legal Advice, Liability, Consumer Protection, Legal Ethics, Technology Law
INTRODUCTION
Artificial Intelligence (AI) has revolutionized the legal field by introducing new ways to research, document, draft contracts and summarize cases. The large language models have become essential tools for lawyers, law firms, businesses and even individuals seeking legal advice. AI technologies offer the opportunity to have better access to legal information and reduce costs associated with legal services. However, there is a twist of AI hallucination. An AI hallucination is a situation when an AI generated response appears to be correct but is false or fabricated. In other words, the information provided by the language model may be fake or contain elements of misinformation. Fake case laws, statutes, citations and legal advice may undermine the credibility of AI technologies. In addition, incorrect data may lead to adverse court decisions, financial losses, professional negligence claims and a general decline in trust in the legal system. The problem of hallucination became apparent when some lawyers used false case laws created by AI in court documents believing them to be real. This controversy has raised many questions about who is responsible for these AI generated fakes and whether companies developing AI software or those using it should be held legally accountable. The ability of AI to generate realistic but potentially dangerous information has prompted a new legal and ethical debates about liability.
UNDERSTANDING AI HALLUCINATIONS
Artificial Intelligence (AI) is the technology which makes computers and machines able to mimic the human processes of learning, understanding, perception, reasoning, thinking, decision making, creating and behaving in an autonomous way. These applications and devices can see and identify objects, understand and respond to spoken languages, continuously learn from experience, give advice on specific topics in detail or even control themselves and perform tasks without human interaction and intellectual power (for example a driverless car).
A generative model is a machine learning model that learns to generate new data similar to the training data. Generative artificial intelligence models learn the patterns and distributions of the training data and then use that knowledge to generate new content in response to new input data.
.jpeg)
