Remedying the Defamatory Bot: The Case for Artificial Intelligence Legislation

Abstract

Artificial intelligence (AI) as a means of content generation has become increasingly common and is being integrated unprecedently into our day-to-day lives. However, with its benefits it also gives rise to the potential for AI-related harm. One example of this is the instance of AI tools producing defamatory material. Complaints of AI making false claims about people have started to emerge and will likely make their way to the courts soon. Though a consideration of previous defamation cases dealing with emerging technologies suggests that the courts may struggle to determine the most appropriate outcome. Especially as the non-human nature of AI speech brings its own unique challenges that may be difficult to harmonise with traditional justifications for defamation liability. On this basis, as well as a consideration of the broader context of the current development of AI, this paper argues that legislation is a more appropriate remedy to defamation caused by AI. Legislation would look to article what developers of AI can reasonably expected to control in relation to their tool and assign responsibility in accordance with this. This approach would better balance the interests at play in disputes concerning AI-related harm; the benefits of developing and having access to AI versus ensuring that its integration does not cause more harm than good. This ensures clarity regarding the relationship between AI and society more generally, something which cannot be achieved effectively through defamation law alone.