Ethical AI
AA: One thing that I’ve been increasingly asked to speak and write about recently has been the whole topic of ethical and responsible AI – it’s becoming big. A lot has been written – there are a lot of policies and frameworks around it. How do you actually make it practical and put it into practice, especially within the APS? What do you think are some of the steps that need to be taken to realize it?
MM: Step number 1 is understanding the “why” of the problem. In the private sector, it might be profit, but in government, the question is, “Why does this problem actually matter? What are the risks to government, to the community, to the nation as a whole, and globally, if we do not solve this problem?” Then, you need to understand what data is required to solve the problem and what risks are associated with the data. People jump on ethics straight away, but it’s not the only risk. I prefer to talk about trust...