Australia’s artificial intelligence (AI) policy debate is at a crossroads. Decisions made in the coming months will determine whether Australians are able to trust AI and capture its benefits or whether we are left exposed to unacceptable and avoidable risk.
Calls to wait for yet further reviews ignore the evidence that action is needed now. Australia cannot afford to repeat mistakes made with previous technological developments, such as social media, where delayed action left communities exposed and regulators playing catch-up.
Global Shield Australia has prepared the attached primer setting out Ten Reasons Australia Needs an AI Act. Now is the time for the government to lead and deliver the safeguards that Australians expect and need to fully capture the AI dividend.
TEN REASONS AUSTRALIA NEEDS AN AI ACT
An AI Actis the best way to:
1. Address the unique hazards posed by advanced AI models, including their demonstrated capabilities to deceive, self-replicate, and pursue their own goals.
2. Establish monitoring and incident reporting requirements for AI across all sectors, to track and respond to harms and better understand its systemic risk.
3. Require AI model developers to take measures to preventtheir products causing harm in all use cases and be transparent to users and regulators.
4. Enable a consistent and certain approach to AI regulation across the fragmented regulatory environment that currently applies to its use.
5. Mandate content provenance and labelling at scale by regulating generative models, a key tool to deal with harms ranging from deepfakes to disinformation.
6. Ensure a consistent approach is taken to assigning legal responsibility for AI actions, including between foreign and local developers, deployers, and end users.
7. Putin place specific security requirements and standards to prevent the misuse of advanced AI models by rogue actors seeking to undermine our national security.
8. Deliver uniform assessment and certification of AI models and tools, making it easier for small businesses and users to use and trust the technology.
9. Align Australia with global partners, minimising compliance friction and enabling Australian businesses to access global AI assurance opportunities.
10.Ensure dedicated regulatory oversight of AI, building capable regulators that can set, adapt and implement baseline standards and coordinate across sectors.
