Watch / Artificial Intelligence (AI)
Keeping humans in the loop as AI faces accountability test
Keeping humans in the loop as AI faces accountability test
About this video
Key points:
South Australia’s AI Royal Commission is examining AI’s impact on productivity, jobs, economic growth, regulation and society, with its final report due in July next year
The inquiry is focused on the forward-looking risks of increasingly autonomous AI systems, including where human accountability should remain
Rocky Scopelliti says AI governance needs safeguards around responsibility, privacy and decisions that should always remain meaningfully human
Rocky Scopelliti, futurologist and adjunct professor at UTS, says the Royal Commission goes beyond AI safety to examine how Australia can use the technology to remain competitive while managing its social and economic risks. Its terms of reference cover productivity, innovation, employment, education, economic growth, social benefits and regulation across government, industry, research and the community.
Scopelliti says the inquiry is unusual because Royal Commissions have traditionally examined what went wrong in the past, while this one is looking ahead at how AI will be deployed and adopted. He argues the process could provide greater certainty around the safeguards and testing regimes needed as AI becomes increasingly embedded in society.
A key issue for Scopelliti is accountability as AI moves from advising people to taking actions on their behalf. He says responsibility cannot be delegated to an algorithm, while future privacy debates will increasingly involve how machines perceive, interpret and infer information about individuals. He also argues society needs to decide which decisions should always remain meaningfully human.