On 31 January 2024, ASIC Chair Joe Longo outlined ASIC’s position on the AI regulatory landscape in Australia. While current legislation provides some regulation, it is insufficient to effectively address the risks associated with AI.
Mr Longo acknowledged that Australia already has general laws that apply to AI, including on privacy, online safety, enterprise, intellectual property and anti-discrimination. He also pointed out that directors’ duties under the Companies Act also apply to companies that use AI.
Additionally, ASIC is already scrutinizing the use of AI in financial services, alongside the recent focus on cybersecurity compliance by AFSL holders.
However, ASIC believes there are still gaps in Australia’s AI regulatory environment. One of the main concerns with ASICs is algorithmic bias that perpetuates discrimination when AI is used in the decision-making process. These biases can manifest themselves in credit scoring, for example, where AI processes can favor certain groups over others based on limited training datasets.
A major challenge for ASIC and other regulators looking to regulate AI is the opacity of the technology (at least in its current form), which makes it difficult to identify and address algorithmic issues. It has become.
Furthermore, the rapid adoption and development of AI requires agile and rapid regulatory responses. AI experts advocate increasing transparency and accountability in AI decision-making by introducing regulatory frameworks that prioritize openness, fairness, and ethical standards.
Longo believes that AI requires more regulation, but also thinks that ASICs could be useful in maintaining the safety and integrity of the financial system for the benefit of consumers.
