AUSTRALIA | As artificial intelligence becomes more ingrained into working lives, questions of accountability loom large. Department of Business Law and Taxation lecturer Dr Estelle Wallingford's cutting-edge research explores who should be held responsible when AI causes harm.
Dr Estelle Wallingford has always had an interest in technology. In secondary school, her teachers encouraged her to consider a career in engineering. Instead, she chose the legal profession - a decision that ultimately landed her at the intersection of two rapidly evolving fields: artificial intelligence (AI) and the law. As a practising lawyer, she gained exposure to the evolving role of AI in professional settings.
Driven by a scholarly interest in the legal and ethical implications of AI she pivoted to academia, where she could explore these complex questions in depth.
"Technology is always changing and evolving, presenting new challenges and new iterations of old challenges that must be addressed," said Dr Wallingford.
"For an academic mind, it's a fascinating field to engage with and contribute to."
'We cannot simply set and forget'
Dr Wallingford's research tackles some of technology's biggest legal questions such as: who is responsible when AI causes harm?
Her recent doctoral thesis, Assigning Liability in the Context of Modern Artificial Intelligence, explores frameworks for determining legal responsibility in AI-related private law disputes.
"Historically, the law has been slow - reactionary rather than proactive - but with AI, we have a technology evolving far faster than our ability to regulate it," said Dr Wallingford.
"We need to attempt to be proactive in our approach, rather than reactive. That is the ongoing challenge - to find the right balance."
Her research proposes a three-phase approach to AI liability, considering factors such as an AI system's level of autonomy and human oversight, advocating for governance frameworks that are both flexible and forward-thinking.
"There's a tendency to oversimplify AI as being one thing, and that's short-sighted - there are multiple types of AI," says Dr Wallingford.
"I emphasise the need for being nimble and flexible, while at the same time having parameters for what is acceptable and what is not."
She believes the same mindset should apply to AI in education.
"AI is rapidly changing all the time. I do really feel that as academics, we need to evolve alongside it," said Dr Wallingford.
"We cannot simply 'set and forget'."
Thinking beyond the algorithm
Despite the interest in AI's ability to boost efficiency, Dr Wallingford warns her students it is a tool that can support – but not replace – critical analysis.
"I teach my students that AI is not a shortcut to thinking. Yes, it can complement what you're doing, but it's only a complement," said Dr Wallingford.
"Even if an AI system can generate something, you need the skills to interrogate its veracity and determine its accuracy."
This is especially critical in disciplines such as law, where ethical and responsible AI use has real-world consequences.
She points to the recent case of an Australian lawyer sanctioned for using ChatGPT to submit false court documents as an example of the risks.
"These sorts of incidents highlight the importance of instilling the ethical and responsible use of AI as early as possible in an individual's career - and this starts at university," said Dr Wallingford.
"Students should have the opportunity to test the parameters of AI in a learning environment before they enter the workforce, where the ramifications of inappropriate AI use are high."
More news here.
