AI Regulations: EU’s Landmark AI Act and Australia’s Guardrails

According to DJ Piper Technology’s Legal Edge blog, global companies adopting artificial intelligence need to understand international AI regulations. The European Union and Australia have taken different approaches, with the European Union enacting the landmark AI Act and Australia adopting a phased approach that uses guardrails.  

The European Union’s AI Act, effective August 2024, sets the most comprehensive precedent worldwide. It categorizes AI systems by risk level, from prohibited practices to minimal-risk applications, each with tailored requirements. High-risk systems are subject to rigorous obligations in governance, documentation, and human oversight, with violations carrying penalties of up to €35 million or 7% of worldwide annual turnover. The AI Act also introduces rules for general-purpose AI models, distinguishing those with systemic risk from those without.

By contrast, Australia is proceeding with a phased, less prescriptive approach. While no binding laws exist, proposed guardrails for high-risk AI emphasize the importance of data governance, transparency, human oversight, and regular testing and monitoring. Australia’s Voluntary AI Safety Standard provides organizations with guidelines for responsibly developing and deploying AI systems, applicable to AI systems at any level of risk.

The difference between the EU’s landmark AI Act and Australia’s guardrails exemplifies regional nuances that demand careful navigation. For global companies, these divergent approaches highlight the need for coordinated strategies. Mastering AI regulations has become a competitive differentiator as much as a compliance necessity, shaping the future of global tech operations.