Review of AI and the Australian Consumer Law

Background

In October 2024, the Australian Treasury launched a consultation on the application of the Australian Consumer Law (ACL) to AI-enabled goods and services. This review forms part of the Government's broader work to clarify and strengthen existing laws to address AI-related risks while supporting responsible innovation.

The review examined questions around consumer protection in an AI-enabled economy, including:

  • How well the ACL protects consumers using AI-enabled products and services
  • Whether current consumer guarantees and remedies are appropriate for AI systems
  • How liability should be distributed between AI manufacturers and suppliers
  • Whether the unique characteristics of AI systems create gaps in consumer protection

Our submission

Good Ancestors' submission argued that the current mismatch between AI capabilities and AI safety presents foundational challenges for consumer law concepts. We emphasised that AI developers should be required to internalise the risks of harm from their systems, rather than shifting accountability to Australian businesses deploying AI who often lack the tools to manage risks effectively.

Our key recommendations were:

  1. The specifics of generative AI behaviours, capabilities and supply chains mean that a technologically neutral approach may not provide practical consumer protections. Policymakers must engage with the unique regulatory challenges that emerge from the specific nature of AI technology.
  2. Policymakers should not focus on the adequacy of regulations for today's AI systems, but forecast forward to AI systems we should expect in coming years.
  3. Australian AI deployers and users will often have limited ability to understand and control the risks of AI systems. Effective regulation must ensure that AI model and system developers are incentivised to take safety engineering seriously.
  4. The potential consequence of an AI accident greatly exceeds the potential consequences from typical consumer goods. The application of consumer law to AI must account for this difference in risk.
  5. Australia should establish an AI Safety Institute with functions relating to understanding technical AI safety, communicating technical information to Government, and facilitating cooperation between Government, academia and industry.
  6. Given practical pressure on businesses to adopt AI even when they cannot assess and mitigate risks, an effective regulatory regime must be able to identify risky behaviour in most cases and impose sufficient consequences to discourage it practically.
  7. Australian consumer law should contribute to global AI safety norm-building by ensuring that those best able to mitigate risks are incentivised to take meaningful action.
  8. Australian consumer law should contribute to global AI safety norm-building by encouraging AI developers to invest in safety. AI models and systems should meet safety expectations consistent with other advanced technology, like aviation.

We also recommended that Model Cards and Responsible Scaling Policies could provide a roadmap for adapting consumer law to AI systems.

The consultation closed on 12 November 2024. Treasury is expected to release its findings in early 2025.