Generative AI and public Interest Journalism
Start year: 2024
Summary: Media organisations are beginning to grapple with the implications of generative AI on news production, and the accelerating impact it may have on mis and disinformation. For at least a decade the prospect of large language models becoming operational and able to produce cogent news output has been considered in the realm of sci-fi. In November 2022, the manufacturer OpenAI changed the discussion, releasing ChatGPT which incorporates a chatbox user interface and has the ability to generate often appropriate and convincing responses to user prompts. Newsrooms around Australia are now investigating its opportunities and challenges. Some, such as the ABC are moving into the development phase for a bespoke ABC LLM. Others are negotiating with manufacturers to sell their output for training purposes. Our research is investigating whether such moves will serve only to further pollute the information eco system and whether a more viable option is to create a coherent set of principles which govern news media usage of generative AI in news output, which also has the capacity to discern when it is appropriate for media businesses to offer content for the training of LLM's.