Editorial responsibility and AI
The fundamental principle of editorial responsibility is not changed by AI — publishers are responsible for the accuracy of everything they publish, regardless of how it was produced. AI-generated content that contains false statements of fact exposes the publisher to defamation liability in the same way as human-produced content with the same false statements. AI origin is not a defence to defamation; it may be relevant to damages (affecting whether the court finds malice or recklessness), but it does not change the fundamental liability analysis.
The practical implication for media organisations using AI in content production: all AI-generated or AI-assisted content must go through the same editorial review process as human-produced content. AI research outputs must be verified. AI drafts must be reviewed for factual accuracy, including specific verification of any quotes, sources, or statistics the AI includes. AI-generated summaries of long documents must be checked against the originals. The editorial responsibility for published content is not dischargeable by pointing to an AI tool that produced or contributed to the content.
Disclosure of AI in journalism
The disclosure question — whether and how to disclose AI use in published content — has been actively debated by media organisations and press standards bodies. The emerging consensus: disclosure is required when AI has been substantially involved in the creation of published content, when readers might reasonably expect the content to be human-authored, and when the use of AI is material to how the content should be evaluated. The specific disclosure requirement varies: some organisations require a byline disclosure ("This article was written with AI assistance"), others a standing disclosure in their transparency documentation, and others case-by-case disclosure for specific types of AI use. ACMA and equivalent regulators are watching media AI disclosure practices and may develop specific requirements.
