Australians say judging when to use AI matters more than knowing how it works
People in Australia are finding AI increasingly hard to avoid, but they rate the ability to question, assess and decide when to use it above understanding how AI systems work, according to a new Community Panel on AI. The panel was hosted by researchers at the ARC Centre of Excellence for Automated Decision-Making and Society (ADM+S) at Swinburne University of Technology, in collaboration with Telstra.
- A Community Panel on AI of 25 Australians aged 19 to 68 was hosted by ADM+S at Swinburne University of Technology with Telstra.
- Two thirds of participants felt pressure to use or avoid AI, and participants ranked critical thinking and judgement above technical knowledge.
- The researchers call the ability to decide when, how or whether to use AI 'AI discernment'.
- Almost all want public AI learning events, but nearly half were unsure where to find local ones.
- The panel is the first stage of a broader national study on AI capability in Australia.
What the panel did
The panel brought together 25 participants aged 19 to 68 from diverse backgrounds and life circumstances. Through workshops and written responses, they described how AI is entering their lives and what helps or hinders their ability to use it effectively and to navigate AI-mediated environments safely.
According to the ADM+S Centre, the panel is the first stage of a broader national study of what it means to be AI capable in Australia. Its insights will inform research into what AI capability looks like in everyday life, and the study is intended to help policymakers, educators and community organisations better target support, learning opportunities and resources.
Pressure to use AI, and the idea of 'AI discernment'
Participants said AI is increasingly difficult to avoid across work, education and everyday life. Two thirds felt pressure to either use or avoid it. The pressure can come through workplace adoption, educational settings, websites using chatbot assistants, or the transformation of search through AI summaries. Two thirds also reported having had disagreements with others about AI use.
Rather than seeing technical understanding as the key skill, participants named critical thinking and judgement as the most important capabilities. The researchers call this 'AI discernment', defined as the ability to critically assess when, how or whether to use AI. That includes knowing when to question or reject its outputs. As one participant put it: "you could be really good at using AI, but you still need judgment and common sense."
Most participants distinguished between using AI to find information or explain things and relying on it for personal advice. Only some are comfortable asking chatbots about health, finance or other personal matters, including parenting tips.
Reliability, deception and trust online
Over half of participants said AI is making it easier to find information online. However, judging the reliability of AI-generated answers and summaries has become a critical new capability, made harder by the unpredictable tendency of AI-generated summaries to contain errors, according to the researchers.
Participants also expressed strong concern about AI being used to deceive or impersonate people. Most said they were no longer confident they could tell whether an image, video or audio had been created or altered with AI. For many, this is changing how they use social media and how they engage with websites or emails from unfamiliar people.
Learning opportunities and accountability
Almost all participants are interested in attending public AI learning events and activities, but nearly half were unsure where to find local opportunities. Most rely on informal sources such as friends, colleagues, social media and AI itself. Many said they depend on people around them for information and support, and those outside work or education reported the least adequate access to learning.
Most participants believe there is not enough public input into AI. They showed strong interest in discussing it with people outside their usual networks and "echo chambers", particularly through local and public forums. They also called for policymakers and organisations to take greater responsibility for AI safety, and for everyday people's voices to be included in decisions about AI.
Dr Dominique Carlon, panel lead and ADM+S Research Fellow, said: "The good news is that people in Australia are already approaching AI with care, curiosity and discernment." Telstra's Acting Chief Sustainability Officer, John Wilkinson, said people are looking for practical support and trusted information as AI becomes more a part of everyday life.
What it means
The findings point to a demand for public education and consultation that goes beyond technical training. Professor Anthony McCosker, project lead and Chief Investigator at ADM+S, said the aim is to ensure these experiences help shape policies and programs that support people through the technological transformation. The project says it aims to make sure policy, education and support initiatives reflect the realities of how Australians engage with AI today.
ARC Centre of Excellence for ADM+S: Knowing when and when not to use AI is a critical new skill, new Australian community panel finds
