The Australian Human Rights Commission (AHRC) recently expressed its opinion on artificial intelligence (AI) decision-making. Consequently, the AHRC has now published its final report detailing AI regulation.
The report followed a three-year investigation into technological growth and AI adoption in Australia. In its report, the AHRC dissected the benefits and risks of AI informed decision-making and provided a slue of recommendations to appropriately extend AI as tools to human decision-making.
The report is divided into the following four categories:
A. The scope and approach to new and emerging technologies:
· This category targets the human rights implications in going forward with AI-based decision-making.
· This involves promoting fairness, equality and accountability in the use and development of AI technology.
· Additionally, by having a concrete strategy in place, the AHRC notes the importance of proper regulation, investment, and education to foster public trust in AI.
B. AI informed decision-making:
· The AHRC provides that AI-based decision-making will need to be appropriately broken down and defined. In the report, this involves looking at both the outcome of a decision and how such a decision was reached. In this way, human rights can be integrated throughout the entire process and offer a level of effective transparency.
· Such decision making also requires a materiality threshold to emphasise that the decision was made substantially by AI and that these generated decisions carry legal weight as means of accountability.
C. Effective regulation:
· The AHRC has emphasised the importance of the lawful, ethical, and transparent development of AI technology.
· Thus, to aid governmental and private bodies developing this technology, the AHRC recommends the establishment of an AI Safety Commissioner to oversee AI implementation, address problems, support regulators, policymakers, governments, and businesses in applying laws and standards in respect of AI-informed decision-making.
D. Accessibility of this technology:
· AI technology must be designed in a manner where it is accessible to all Australian people.
· The last category of the report highlights the necessity to abolish disability discrimination and create technology that will benefit people regardless of any disabilities.
The AHRC’s final report also covers 38 recommendations for how the governmental and private sectors should act and operate when dealing with AI technology. Most notably, this boils down to liability, transparency, and the human rights impacts of using AI-based decision-making. By auditing internal AI use, these bodies are expected to implement policy and safeguard measures to ensure the appropriate use of AI technology. Thus, the report emphasises a proactive approach to prepare the Australian society for wider AI adoption going into the future.