Socially responsible insurance in the age of artificial intelligence

Primary Researchers: Professor Seth Lazar and Professor Kimberlee Weatherall

Institutions: The Australian National University and the University of Sydney.

As artificial intelligence (AI) increasingly becomes part of our lives, the debate about its potential for both positive and negative impacts has come to the fore in Australia. ANU Professor of Philosophy Seth Lazar is leading a multidisciplinary team researching how to maximise the potential benefits of using AI as a tool in insurance while also ensuring that it avoids unintended social costs. 

Professor Lazar’s ARC Linkage Project acknowledges that AI enables people to make inexpensive, reliable predictions, and is therefore a powerful tool to help society manage risk more efficiently. At the heart of the research is the concern that without appropriate safeguards the use of AI in insurance carries considerable risks. 

‘The appeal of using machine learning to enable more individualised risk predictions is undeniable, but without careful attention it is likely to lead to new forms of discrimination and troubling privacy violations, and potentially threaten the social function of insurance,’ Professor Lazar said. 

Why insurance? In a presentation on the project at the 2023 International Congress of Actuaries, Professor Lazar’s collaborators, Chris Dolman and Kimberlee Weatherall, noted that insurance was ‘not just another industry’. It is treated differently in policy and law because of the vital role it plays in society – for underwriting homes, cars, workers compensation, and public liability, for example. 

The project aims to design practical interventions – responsible design workshops, practical guidance, regulatory proposals, new algorithmic tools – that realise the benefits of using AI while mitigating the costs. Ultimately, the researchers believe AI and machine learning can be deployed to promote and even enhance the role insurance plays in society. 

‘We are at a critical juncture in the development and societal adoption of AI, as major progress in Machine Learning research is leading to widespread adoption throughout the economy,’ Professor Lazar said. 

‘Only by situating these new technologies in their social context will we be able to safely harness those opportunities – as well as responsibly decide when AI is not the answer.’ 

Image: iStock