The federal government has released a proposal for mandatory guardrails to be adopted by Australian companies developing or deploying artificial intelligence (AI) in “high-risk settings”.
Through the Department of Industry, Science and Resources, the federal government also released an AI safety standard with very similar guardrails that can be voluntarily applied to any risk setting.
The proposals paper is open for consultation until 4 October 2024 while the voluntary safety standard is already in effect.
AI regulation has been on the government’s agenda since the release of its discussion paper title Safe and responsible AI in Australia in June 2023, which was followed by an interim response paper in January this year.
The draft regulations are set to have wide applications, as such law firm Clayton Utz notes Australian businesses must start understanding their implications and evaluate approaches to manage any AI risks.
The Business Council of Australia echoes this sentiment. The council supports a risk-based approach to AI which implements appropriate regulatory settings while ensuring protections without stymieing growth.
Business Council CEO Bran Black says Australia must therefore adopt an AI framework that safeguards the community, while encouraging businesses to use the technology for its benefits such as productivity.
Productivity is at its lowest point in decades in the country and AI is poised to provide a major opportunity for workers and businesses to bolster output and share in the resulting economic benefits, the CEO says.
“Key to the successful integration of AI into the Australian economy will be striking the right balance between protecting individual rights while also ensuring our nation is best placed to reap the productivity benefits AI will bring,” Black says.
Considerations of new regulations must ensure there is no duplication of existing laws, which will make it harder for businesses of all sizes to comply, the CEO notes.
The WA Data Science Innovation Hub believes using data to solve complex business and societal challenges is fast emerging as a key advantage for many territories globally. The vast array of data collected by governments and businesses provides a chance for all states and territories to develop innovative solutions to address challenges and capture opportunities, but requires requisite skills and collaboration, according to the Hub.
many of Australia’s existing technology-neutral laws and regulations already apply to the use of AI technologies. This includes regulatory frameworks regarding consumer protection, privacy, anti-discrimination, as well as sector and profession specific requirements.
The Australian Government believes mandatory regulation is needed because self-regulation is not sufficient as the implementation of AI will likely highlight gaps in these regulations.
Keeping trust and confidence on track
Clayton Utz suggests AI regulation is also needed as public trust and confidence of the technology in Australia remains low.
“An effective system of AI regulation that mandates the responsible use of AI should improve public trust and encourage AI adoption. However, the government has concluded that our current regulation of AI is not fit for purpose (based on its consultation process),” the law firm says.
“Australia has no AI-specific regulation. We do have many existing laws that regulate AI technology in a non-specific way such as privacy, intellectual property, competition and consumer protection, corporations law, online safety, anti-discrimination, criminal and sector specific laws.
“But there are gaps and uncertainties, such as unclear accountability for actions taken by AI, insufficient obligations for transparency about AI systems, and gaps in enforcement measures.
“Also, many of the existing laws focus more on liability for causing harm rather than on preventing the harm by imposing ‘ex ante’ (before the event) risk management obligations on those in the AI supply chain developing or deploying the technology.”
According to Minister for Industry and Science Ed Husic, the Responsible AI Index 2024 shows Australian organisations often overestimate their capability to employ responsible AI practices. While 78% of businesses reportedly believed they were implementing AI safely and responsibly, only 29% of the time this was found to be the case.
The index found on average, Australian businesses are adopting just 12 out of 38 responsible AI practices.
Adoption of AI is emerging as a game-changer for the resources sector, enabling more efficient exploration through automation, while generating greater yields, maximising extraction, and improving safety and operational performance.
Mining giant Newcrest (ASX:NCM) was an early adopter of AI when in 2019 the company collaborated with Insight Enterprises (NASDAQ:NSIT) to deploy Microsoft cloud, AI, and IoT technologies at its Cadia Valley gold mine in New South Wales. The technology was used at the underground mine to monitor and manage crushed ore levels as Newcrest endured challenges managing the levels in the site’s bins.
AI-powered systems are now being rolled out across mining operations broadly to enhance resource estimation accuracy. By examining geological data patterns and incorporating historical mining data, AI algorithms are providing more precise estimates of mineral reserves.
Companies are also improving efficiency and safety in mining operations. For example, AI-driven predictive maintenance systems can monitor equipment performance and proactively help maintenance scheduling, maximising workflows and extending the lifespan of assets.
In terms of safety, human error and workplace incidents are being reduced as autonomous vehicles and machinery equipped with AI algorithms can navigate remote complex terrains, optimise site routes, and perform a range of tasks with precision.
According to Pitchbook data, investment in AI model safety has gained momentum as a growing number of companies seek to address risks such as deep fakes and model hallucination.
Conundrum Australia is one such company in the space focusing on creating advanced machine learning software for the digital transformation of the resources sector. The company’s mission is to expedite the shift of the resources industry to the “border of the physical limit of manufacturing efficiency”.
One example of Conundrum Australia’s AI solutions is optimising the crushing circuit to achieve maximum performance while realising required product quality. The Conundrum system works with all types of crushing circuits to improve efficiencies even with incomplete mounting of sensors.
Earth AI deploys precise mineral detection technology to move from detecting a prospect to drill-testing in just three to six months. The company focuses on greenfield exploration in Australia and only partners with publicly listed resource companies.
Earth AI has built the Mineral Discovery Engine – a system of key processes and technologies that bring precision, efficiency, and scalability to critical minerals exploration. Its AI platform “repeatedly discovers” new and overlooked prospects for a multitude of commodities.
In an indication of the burgeoning sector, AI safety startup Safe Superintelligence (SSI) raised US$1 billion to acquire computing power and build a team of researchers and engineers, as reported by Mining.com.au last week.
The deal reportedly gives SSI a US$5 billion valuation with participants in the raise including Andreessen Horowitz, Sequoia, DST Global, SV Angel, and NFDG.
Co-founded by former OpenAI researchers Daniel Levy and Ilya Sutskever along with angel investor Daniel Gross, SSI is starting the world’s first straight-shot SSI lab, with one goal and one product – called Safe Superintelligence.
Write to Adam Orlando at Mining.com.au
Images: Stock, Insight & Earth AI
