Trust in Artificial Intelligence: Australian Insights 2020

Navigate between the chapters below for topic-focused insights and data.

Trust is central to the acceptance of AI, and is influenced by four key drivers

Our results confirm that trust strongly influences AI acceptance. Of these drivers, the perceived adequacy of current regulations and laws is clearly the strongest. This demonstrates the importance of developing adequate regulatory and legal mechanisms that people believe protect them from the risks associated with AI use.

Australians have low trust in AI systems but generally ‘accept’ or ‘tolerate’ AI

Trust in AI systems is low in Australia, with only one in three Australians reporting that they are willing to trust AI systems. Almost half of the public (45%) are unwilling to share their information or data with an AI system and two in five (40%) are unwilling to trust the output of an AI system (e.g. a recommendation or decision). Many Australians are not convinced AI systems are trustworthy. However, they are more likely to perceive AI systems as competent than designed to operate with integrity and humanity.

This national survey is designed to understand and quantify Australians’ trust in and support of AI, and to benchmark these attitudes over time.

AI is used in a range of applications, such as calculating the best travel route to take in real-time, predicting what customers will buy, identifying credit card fraud, helping diagnose disease, identifying people from photos, and enabling self-driving vehicles.

By taking the first deep dive into the question of trust, this research provides a comprehensive and nuanced understanding of Australians’ overall trust in AI systems, as well as in specific AI applications in the domains of healthcare, policing, HR and financial investment. These domains represent common applications of AI that relate to citizens, employees and consumers

This research provides insights into the key drivers of trust, community expectations and confidence in the regulation of AI, expectations of the management of societal challenges associated with AI, as well as Australians’ current understanding and awareness of AI. Importantly, the findings provide a clear understanding of the practices and principles Australians expect organisations to use to responsibly develop and ethically deploy AI in society and the workplace.

These insights are relevant for building and maintaining the trust and acceptance of AI system by the Australian public, as well as informing policy and practice across government, business and non-profits.

Artificial Intelligence (AI) refers to computer systems that can perform tasks or make predictions, recommendations or decisions that usually require human intelligence. AI systems can perform these tasks and make these decisions based on objectives set by humans but without explicit human instructions.

We asked Australians how much they trust, accept and support AI in general, as well as specific applications of AI.

Australians are ambivalent about trusting AI although they generally accept or tolerate AI. Public support for the development and use of AI depends on its purpose, with most support for its use in healthcare (e.g. disease diagnostic) and least support of its use in human resources.

We asked how much confidence Australians have in different entities to develop and use AI, as well as regulate and govern AI.

We found that the public are most confident in Australian research and defence organisations to develop and use, and regulate and govern AI, and least confident in commercial organisations.

We explored the expectations the public have around AI development and regulation, including the extent to which they think regulation is necessary, who should regulate, and whether current regulations and institutional safeguards are sufficient.

We found that regulation is clearly required and Australians expect external, independent oversight, yet few perceive current regulations to be sufficient.

Eight AI design and governance principles and practices are highly important for trust.

These principles are: technical robustness and safety; data privacy, security and governance; human agency and oversight; transparency and explainability; fairness and non-discrimination; accountability and contestability; AI literacy and risk and impact mitigation.

AI is becoming more common in the workplace yet Australians vary in their use of AI at work.

Most report that little or none of their work involves AI, but given many Australians have a low understanding and awareness of AI use, these figures may reflect that AI is not being used at work, or that people are not aware of its use at work. Australians are generally comfortable with AI use at work when it is not focused on them. Most Australians don’t believe AI will create more jobs than it will eliminate, and if jobs are automated, workers will expect support such as being redeployed or retrained.

The pervasive use of AI in society is leading to a range of challenges.

This survey found that Australians expect all AI challenges to be carefully managed, with a particular focus on how data will impact people in the near future.

To identify how well Australians understand AI, we asked about AI awareness, knowledge about AI and interest to learn more.

In general, Australians have low awareness and understanding of AI and low knowledge of its use in common everyday applications, yet most want to know more about AI. Results in Chapter Eight show that awareness and understanding of AI influences trust in AI systems.

To identify the most important drivers of trust and acceptance of AI systems, we used an advanced statistical technique called path analysis.

These findings revealed that trust is central in AI acceptance, and that the strongest driver of trust is believing that current regulations and laws are sufficient to ensure AI use is safe.

University of Queensland Researchers
Professor Nicole Gillespie, Dr Steve Lockey and Dr Caitlin Curtis

KPMG Advisors
James Mabbott, Richard Boele, Ali Akbari, Rossana Bianchi and Rita Fentener van Vlissingen

Acknowledgements
We are grateful for the insightful input, expertise and feedback provided by members
of the Trust, Ethics and Governance Alliance at The University of Queensland, particularly Dr Ida Someh, Associate Professor Martin Edwards and Professor Matthew Hornsey, KPMG Partners Phillip Sands, Scott Guse, Joel Di Chiara and Leo Zhang, as well as domain expert input by Greg Dober, Mike Richmond and Professor Monica Janda.

Trust in Artificial Intelligence: Australian Insights 2020
The report explores Australians’ attitudes towards AI, and how trust can be strengthened to realise its full potential. Download PDF (4.5MB)