Large language models (LLMs) are a type of language model that form the basis of certain generative AI systems, a subset of ‘artificial intelligence’ or AI.

New services powered by LLMs have significant potential to benefit consumers and boost productivity. This technology can be applied in various ways to enhance the quality of goods and services for consumers and improve efficiency and productivity in sectors such as programming, legal and support services, education, academia, media and public relations. It also has the potential to enhance digital literacy and create positive online experiences for all Australians.

However, the use of these models carries risks, including increased exposure to harmful content such as misinformation, abusive or exploitative content, manipulation, coercion, scams and fake reviews. There are also concerns related to anti-competitive behaviour and threats to privacy and safety due to the collection and use of personal data.

While LLMs will likely intersect with a wide range of legal and regulatory issues, such as copyright and anti-discrimination laws, this paper specifically addresses how LLMs intersect with the regulatory responsibilities of DP-REG members, which include consumer protection, competition, the media and information environment, privacy and online safety.