Key facts
- Australia is introducing new national rules for government use of AI in automated decision-making.
- The rules will focus on fairness, accuracy, and transparency.
- Legislation for a digital duty of care for AI companies is underway.
- Privacy laws will be reformed to modernize personal data protection.
- AI safety in the workplace will be a focus of a new work plan.
- New rules for energy-intensive datacentres will be legislated from next year.
The Australian government is set to implement new national rules governing the use of artificial intelligence in automated decision-making by government departments and agencies. These regulations aim to ensure safety, fairness, accuracy, and transparency in AI processes, extending to consumer protections, workplace safety, and privacy.
Attorney-General Michelle Rowland will lead the development of these rules, which come as major AI companies like Anthropic, OpenAI, Microsoft, and Google are increasing their presence in Australia. The government is also progressing legislation for a digital duty of care, spearheaded by Minister for Communications Anika Wells, to place responsibility on AI companies for building in safety features and addressing potential harm.
Further reforms to privacy laws are planned to modernize and simplify personal data protection. Additionally, a new work plan will address AI safety in the workplace. Prime Minister Anthony Albanese announced plans for a dedicated Office of AI within his department to manage and maximize opportunities from AI.
New regulations for large, energy-intensive datacentres are also slated for legislation from next year. Assistant Minister Andrew Charlton noted that Australia already has over 200 datacentres consuming approximately 2% of the national energy grid, with projections for significant growth. The upcoming standards aim to support local communities, prevent negative impacts on housing and schools, and avoid increasing power prices or consuming drinking water.
However, Deputy Liberal leader Jane Hume criticized the government's approach, stating that regulating AI by focusing on datacentre rules is insufficient and akin to regulating the fishing industry by licensing boat building. She suggested that overregulation could also be problematic, emphasizing the need for certainty while exploring alternative opportunities.