Key points

• The Australian Government’s Department of Industry Science and Resources has released the ‘Safe and Responsible AI’ Discussion Paper, prompting submissions from the public on appropriate regulatory and governance responses to AI in Australia going forward.
• The aim of the Discussion Paper is to request responses which strike a balance between mitigating the potential risks from AI and automated decision-making (ADM), increasing public trust and confidence in their development and use, and promoting innovation in the AI space.
• Submissions close on 26 July 2023.

Introduction

The Australian federal Government released a discussion paper on 1 June providing an overview of existing AI governance in Australia and international developments in relation to AI, and seeking feedback on whether further governance and regulatory responses are needed. The “Safe and Responsible AI in Australia Discussion Paper” (Discussion Paper) was released by the Department of Industry, Science and Resources.

The Government is reviewing its approach to AI governance in light of emerging risks associated with the widespread and rapid implementation of AI around the world.

The Discussion Paper notes that in Australia, there are currently some protections in place in relation to AI which are addressed through existing general regulations (i.e. data protection and privacy laws, Australian Consumer Law and online safety laws) and also sector-specific regulations (i.e. specific laws in relation to therapeutic goods, airline safety and financial services).

The Discussion Paper’s ultimate goal is to obtain feedback from the Australian public on ensuring the Australian Government’s approach to AI governance mitigates risk and fosters public trust in the development and use of the technology. Public consultation closes on 26 July 2023.

Key Benefits and Risks of AI

According to the Discussion Paper, McKinsey estimates that AI could cumulatively add between $1.1 trillion and $4 trillion to the Australian economy by the early 2030s. AI is undergoing a highly innovative and transformative period and has significant potential benefits, such as optimisation of engineering designs to improve building safety, improved analysis of medical images in hospitals and cost-reduction in the provision of legal services. With this great opportunity comes significant risk. The Discussion Paper expresses particular concern regarding:

• algorithmic bias as one of the biggest dangers of AI, citing well-publicised discrimination against individuals based on protected attributes such as race and sex;
• the risks associated with the technical aspects of AI including system accountability and transparency, and the quantity and reliability of data used to train models for their intended purpose;
• anti-competitive market impacts of monopolised ownership of large, rich datasets; and
• lack of transparency (e.g. of algorithms and datasets) inhibiting accountability, risk mitigation, and appropriate allocation of liability.

Policy Goals: Where to Next?

The Discussion Paper canvasses differing legislative and regulatory approaches to governing AI from several jurisdictions around the world. This analysis demonstrates varying levels of regulatory strength internationally, ranging from voluntary standardised self-testing tools for businesses in Singapore to mandatory security review, user verification and registration for certain algorithms in China.

Influenced by the international approaches, the Discussion Paper proposes a draft risk management framework for managing AI risks in Australia, which triages use according to risk level and sets out corresponding obligations (similarly to the EU).

Next Steps: Have Your Say

The Discussion Paper seeks specific feedback on a number of areas including:

• how key concepts in AI should be defined (e.g. ‘machine learning’);
• potential gaps in current regulatory approaches and non-regulatory approaches;
• whether governance measures taken in other jurisdictions are suitable and adaptable for Australia;
• how the Government should target or prioritise its responses;
• trade and infrastructure implications; and
• the utility of risk-based approaches.

The Australian federal Government is seeking feedback on the Discussion Paper and consultation is open until 26 July 2023. Dentons' AI team in Australia can assist you with monitoring legal developments, making submissions and assessing any legal issues with your current activities in AI.

Please contact Matthew Hennessy, Partner, Robyn Chatwood, Partner or other members of the Dentons AI team for further assistance.

The full Discussion Paper can be found at this link.