Ed Husic is missing an opportunity to position Australia as a global leader in regulating artificial intelligence and setting clear guidance for private investment in adoption of AI applications.
Almost eight months after promising sweeping AI regulations amid concerns from tech chiefs that humans face the “risk of extinction” without mitigation, Husic is stuck in a holding pattern.
At British PM Rishi Sunak’s AI Safety Summit on November 3, Husic signed up Australia to the Bletchley Declaration “affirming that AI should be designed, developed, deployed and used in a manner that is safe, human-centric, trustworthy and responsible”.
This is on-brand for the Albanese government, which is quick to sign global pledges but slow to enact pathways to deliver on international promises. The EU is the world leader on AI regulation, passing its stand-alone Artificial Intelligence Act in December. Other countries including Canada and South Korea are well ahead of Australia in developing AI-related legislation, while the US and Britain have developed strong principles-based guidance for both public and private sectors.
Husic’s softly-softly approach on AI comes despite the majority of respondents to the government’s consultation process expressing the need for regulation and legislation to govern AI in Australia. While most tech companies would prefer voluntary codes, many experts believe mandatory rules must be enforced.
Establishing an “expert advisory body to support the development of options for further AI guardrails” is another example of the government kicking the can down the road.
Given the speed in AI advancements, the clock is ticking for Australia to get regulatory frameworks in place and unleash the estimated $170bn to $600bn per year boost to GDP that AI and automation promises.
While co-design and getting the balance right on innovation is important, the government could move faster to establish greater certainty around AI and its applications. Lack of certainty could mean companies and organisations delay AI investments over fears that the use of some technologies could be repealed, potentially holding back Australia’s ambition of becoming a world-leading digital nation.
Instead of action, the government has pledged another round of consultation with industry to develop a voluntary AI Safety Standard implementing risk-based guardrails and to explore voluntary labelling and watermarking of AI-generated materials. The government’s interim response concedes there are public concerns about the safety, testing and operation of AI systems that could ultimately undermine trust in these technologies.
Another major issue is around “training data”, which lies at the heart of AI systems and can insert biases that result in “inaccurate or unfair outcomes for people or groups”.
Accelerating AI adoption in Australia presents a myriad of opportunities but also comes with significant risks.
To harness its potential and mitigate its dangers, AI must be regulated effectively.
Coalition puts UN on notice
The Coalition’s new foreign affairs spokesman David Coleman has signalled a Dutton government would push for reforms of the UN amid criticisms of the body’s stance on Israel.
Confident Albanese immersed in self-belief
The Prime Minister’s self-confidence in his campaigning prowess and popularity remains high, despite Labor being on track to lose majority government and his personal appeal hitting new lows.