EU Artificial Intelligence Act: Implications and opportunities for businesses in Australia and New Zealand
Fujitsu / April 3, 2024
The world’s first legal framework on artificial intelligence (AI), the EU AI Act was approved by Members of the European Parliament in March 2024. This landmark legislation sets the foundation for how European organisations will develop and use AI solutions, ushering in a new era of regulation.
At its core, the AI Act – details of which can you find here – takes a risk-based approach, ranging from outright bans on dangerous AI applications to minimal regulations of very low risk solutions.
What are the challenges with the Act?
While the AI Act represents an important first step, there are many challenges with its implementation. Firstly, the rapid pace of AI innovation will render parts of the Act redundant very quickly, requiring constant updates to address future technological advancements.
Secondly, there are concerns regarding the Act's omission of regulation of AI for military and defence purposes, leaving individual states to navigate this complex terrain independently.
Providers of large language foundation models will also face hurdles in meeting documentation and copyright requirements, as most of their training data includes copyrighted information.
Ambiguities in the Act's language also pose further challenges to implementation and enforcement. For example, it is unclear whether the legislation on foundation models apply to those that already exist, including ChatGPT4.
Despite the implementation challenges, the EU AI Act is still a step in the right direction towards establishing guardrails for the responsible development and deployment of AI systems. As with everything, we will see the laws change and evolve with time.
What does this mean for organisations in Australia and New Zealand?
Like the GDPR regulations that aim to protect the privacy of personal information of European citizens, the EU AI Act will apply to any organisations that engage with the EU markets. And therefore, it is important for companies in Australia and New Zealand that develop or use AI solutions to understand and comply with the Act to avoid substantial fines.
In addition, the Australian and New Zealand governments are likely to closely examine the EU AI Act when defining their own AI legislation. The Australian Government in 2023 requested industry feedback on AI regulation and is expected to pass its own legislation over the next 12 months.
Similarly, the New Zealand government is also likely considering its own AI legislation in the future.
As part of their preparations for similar legislation in Oceania, it is worthwhile for organisations that develop AI solutions to familiarise themselves with the EU AI Act.
What steps can organisations in Australia and New Zealand take to get ready?
Preparing for impending AI legislation requires a proactive approach. Organisations must evaluate the impact of AI on their operations and markets. They should review relevant government policies and industry legislations, as well as their own policies and processes, to determine what governance they need to put in place for their AI solutions. This is essential to mitigate risks and ensure responsible AI deployment.
Ethical AI isn’t just a goal; it is a strategic imperative for long term success. It is crucial that organisations understand and ensure their AI solutions are implemented safely, ethically, and compliant with legislation. By embracing responsible AI practices, organisations will foster trust, transparency, and accountability to help drive AI innovation forward.
Worried about how the new EU legislation on AI will impact your business?
Don't navigate these complex waters alone. Fujitsu's expertise in AI implementation can help your organisation stay compliant while leveraging the power of AI to drive growth. Contact us today at dataandai@fujitsu.com to discuss your AI needs and let us guide you through this new legislative landscape.
