Experts tapped to develop Australia’s AI rules

Avatar photo

Joseph Brookes

Options and thresholds for regulating artificial intelligence in Australia will be developed over the next five months by renowned artificial intelligence expert Toby Walsh, the CSIRO’s chief scientist Bronwyn Fox and intellectual property barrister Angus Lang.

The three experts are part of a 12-person Artificial Intelligence Expert Group to be fully unveiled by Industry and Science minister Ed Husic on Wednesday, with the technical experts expected to be joined by human rights and civil society counterparts.

Update: the full panel has been announced and is listed below.

Mr Husic committed to the panel last month to explore mandatory regulation options for AI development and deployment in high-risk settings as pressure grows to put more safeguards around the technology.

The experts will start with defining ‘high risk’ AI and then look at regulatory options.

UNSW Scientia Professor Toby Walsh has been a strong advocate for regulating the highest risk use of AI. Image: UNSW

“This Artificial Intelligence Expert Group brings the right mix of skills to steer the formation of mandatory guardrails for high-risk AI settings,” Mr Husic said.

“With expertise in law, ethics and technology, I’m confident this group will get the balance right.”

The government has already said where existing frameworks like privacy or consumer law can be applied to AI they will be, but signalled new specific mandatory rules will be needed.

The government’s interim response to consultation on a risk-based approach to safe and responsible AI flagged new regulations could require independent testing of the technology when deployed in high risk settings, as well as disclosure of its use to end users and designated responsibility for AI safety to specific individuals.

“It’s imperative sophisticated models underpinning high risk AI systems are transparent and well tested,” Mr Husic said.

The full 12 person panel is:

  • Professor Bronwyn Fox: CSIRO chief scientist.
  • Aurélie Jacquet: Chair of Australia’s national AI standards committee, OECD expert on AI risks, and advisor on international AI certification initiatives.
  • Dr Terri Janke: An international authority on Indigenous Cultural and Intellectual Property (ICIP).
  • Angus Lang SC: A leading legal practitioner and contributor on intellectual property law and AI.
  • Professor Simon Lucey: Director of the Australian Institute for Machine Learning at the University of Adelaide.
  • Professor Jeannie Paterson: Founding co-director of the Centre for AI and Digital Ethics.
  • Professor Ed Santow: Co-founder of the Human Technology Institute, former Australian Human Rights Commissioner.
  • Professor Nicolas Suzor: A Future Fellow at QUT and chief investigator of the ARC Centre of Excellence for Automated Decision-Making & Society.
  • Professor Toby Walsh: UNSW Scientia Professor and widely recognised voice on AI development, with leading roles at Data 61 and several international fellowships.
  • Professor Kimberlee Weatherall: A chief investigator with the ARC Centre of Excellence for Automated Decision-Making and Society.
  • Professor Peta Wyeth: Researcher on human computer interaction, human-centred artificial intelligence, and design practice and management.
  • Bill Simpson Young: Co-founder and CEO of Gradient Institute and leading technologist for the safe and responsible use of AI.

The interim response also committed to a new voluntary standard for less risky uses of the technology and the introduction of labelling and watermarking requirements for AI-generated materials.

The response confirmed the lowest risk use will continue largely unimpeded by regulation.

Experts welcomed the move to new regulation and a focus on safe and ethical use, but questioned the pace of the Albanese government’s response as other nations raced ahead with AI rules and a lack of public investment in the technology.

UNSW AI expert Professor Toby Walsh, who is part of the new advisory panel, described the government’s interim response last month as “a little light and a little late” compared with efforts in other jurisdictions.

He has previously waned that AI is fundamentally different to other technologies because of it deployment speed and potential impact on people’s lives.

“One of the unique challenges AI throws up is the speed with which it is being developed and adopted.  We’ve never had a technology before that could so quickly reach into our lives,” Professor Walsh said.

“This makes setting the guardrails both difficult and critical.”

Professor Walsh has also warned about the uptake of AI by military users and called for autonomous weapons to be regulated.

The government has said its new AI rules will seek safety without compromising the productivity benefits of the technology — estimated to, along with automation, be a potential $100-to-$600 billion a year add to Australia’s GDP.

Panel member and CSIRO chief scientist Professor Bronwyn Fox is an engineer with a background in advanced manufacturing and translating research into industrial outcomes.

She said AI is already creating enormous benefits and will create even more.

“Australia already has extensive AI capabilities, and our researchers are using AI to tackle everything from improving cybersecurity to delivering better healthcare outcomes, from bushfire management to boosting agricultural productivity,” Professor Fox said.

“Our challenge — and something this expert panel will be looking at closely — is to understand the many benefits AI will create for Australia but also to ensure that it is developed within responsible guardrails and on a strong foundation of ethical principles and governance structures.”

The UNSW and CSIRO professors will be joined on the panel by Angus Lang SC, an intellectual property legal expert who has lectured in Europe. He has been a barrister at Tenth Floor Selborne/Wentworth Chambers for almost twenty years.

The advisory group will be in place until June 30 this year. But with no timeline on the new regulations yet, the government may extend it’s tenure to provide ongoing advice on AI.

This article was updated to include the full 12-person panel.

Do you know more? Contact James Riley via Email.

Leave a Comment