Making the most of the AI opportunity: productivity, regulation and data access
Research paper
Policymakers around the world are considering how to best harness the transformative potential of artificial intelligence (AI) for productivity while trying to anticipate and limit any associated risks. But as AI technology becomes more pervasive and sophisticated, each country is faced with its own unique challenges and opportunities. In this set of three papers, the Productivity Commission places Australia’s AI opportunity in a global context, to consider how our governments can help to maximise the potential economic benefits of this evolving technology. Paper one outlines how Australia stands to benefit most from AI technology, and consequently, where governments should focus their policy efforts. It finds that Australia’s opportunities in the short-term are further down the AI value chain, in adapting general purpose models to more specific, local use cases; or in implementing the technology, particularly through digitised firms and software as a service. Australia could also play a role developing AI models that can be trained on smaller, high-quality datasets. Government should focus on ensuring Australia has the skills and digital infrastructure to integrate AI in these ways and modelling best practice. They should also play a role in building public trust, ensuring trustworthiness, and establishing frameworks for data access and protections. Paper two considers governments’ other major contribution to AI policy: regulation. Here too we must view Australia’s situation from a global perspective. Given our likely role in the AI value chain, Australian regulation will be driven, at least in part, by international approaches. If we set local policy that is out of step with emerging global norms, we risk leading AI suppliers to bypass Australia, harming our economy. Fortunately, many of Australia’s existing technology-neutral laws and regulations already apply to the use of AI technologies. This includes regulatory frameworks in areas such as consumer protection, privacy, antidiscrimination, negligence and sector specific and profession specific requirements. However, the implementation of AI will likely highlight gaps in these regulations. We provide a practical guide to help government respond proportionately and effectively to such regulatory challenges. The challenges presented by AI are particularly acute in the case of data. Data is a vital input into AI technologies and, generally, increased accessibility of quality data would likely contribute to productivity. But at the same time, wider use of data (particularly with inadequate regulation) increases risks to individuals and raises questions about the rights of those generating, curating and using data. Paper three considers how AI raises the stakes for data policy, and what Australian policymakers should do to address the new questions about data rights and incentives that AI presents. One challenge is to improve public confidence in data‑sharing, including through enforcement of existing protections for individuals. Another challenge is to establish clear and consistent arrangements for text and data mining (TDM) for the purposes of training AI models. Australian data remains an underutilised resource and will be increasingly valuable to harness the potential of AI.
We value your comments about this publication and encourage you to provide feedback.
