“This promises a range of benefits—AI could help solve complex problems, improve efficiency across a range of industries, and enable new or improved products and services,” the Australian Communications and Media Authority (ACMA) said when releasing its reports on artificial intelligence and the Internet of Things.
“However, there are also challenges raised by AI that need to be addressed and managed,” the ACMA cautions in the overview to its report.
“There has been a trend to frame these as ‘ethical’ challenges or issues. Some of those most commonly discussed include providing transparency around how AI reaches a decision or prediction, ensuring that the outcomes of AI are fair, ensuring a human is accountable for the outcomes of AI, and for the protection of privacy.
The ACMA says, however, that there are also challenges raised by AI that need to be addressed and managed.
“There has been a trend to frame these as ‘ethical’ challenges or issues.
“Some of those most commonly discussed include providing transparency around how AI reaches a decision or prediction, ensuring that the outcomes of AI are fair, ensuring a human is accountable for the outcomes of AI, and for the protection of privacy.”
The ACMA says that policymakers and regulators are working to address the challenges of AI to support “public trust and guide AI development and use towards beneficial outcomes”, and
“the ACMA has a role to play in facilitating and supporting these efforts at a sectoral level.”
“The regulatory framework administered by the ACMA works alongside a range of other regulatory arrangements, including consumer safeguards, competition and privacy laws,” the ACMA also notes.
“Because of the intersection between these regulatory regimes and public policies, this paper includes discussion of issues that are also relevant to regulators across these areas.
“However, because we are focused on our regulatory framework, we have not included in-depth discussion of some prominent issues relating to AI that sit outside of our remit. These include:
Privacy, data protection and governance: domestic and international privacy laws and data protection laws play a central role in managing risks in AI, given AI technologies are built on data. Parts of the ACMA’s regulatory framework intertwine with Australia’s Privacy Act 1988 (Privacy Act), which is administered by Australia’s privacy regulator. We have not included detailed discussion of the range of issues AI raises for privacy. For further information on these, please refer to relevant work published by the Office of the Australian Information Commissioner (OAIC), state privacy regulators and the Australian Human Rights Commission (AHRC).
AI technologies may have security vulnerabilities that need to be resolved. AI could also potentially be used to conduct cyberattacks. There are laws in placearound cybersecurity (for example,the Criminal Code Act1995)and privacy lawsthat set requirements for the protection of personal information(for example,the Privacy Act).
Competition: concerns have been raised about the behaviour of digital platforms in the supply of advertising or content online, where AI playsa role in display and distribution. Relevantly, the Australian Competition and Consumer Commission (ACCC) is conducting inquiries into the markets for the supply of digital platform services, digital advertising technology services and digital advertising agency services. The ACCC is also developing a mandatory code of conduct to address bargaining power imbalances between Australian news media businesses and each ofGoogle and Facebook.
Product safety and liability: product safety regulation (for example,Australian consumer law) that may be relevant to AI-enabled products sits outside of the ACMA’s remit.