ANZ’s progress in 2023
Last January we were all talking about ChatGPT. In truth many of us initially perceived this as an amazing toy. Writing speeches, learning quantum physics, or replacing Google searches were the shared holiday anecdotes. This swiftly turned to how this could help ANZ in our purpose.
Things moved pretty quickly. We started with a bold position not to block ChatGPT, enabling our teams to explore and learn about this game changing technology.
In March we established an AI Centre of Experiment in Bengaluru to build a virtual team tasked with exploring AI proof of concepts. We connected with major technology partners, ran a hackfest and built a backlog of ideas. One of the capabilities we delivered was our private and safe in-house chatbot, Z-GPT.
Our engineering teams developed an experiment with GitHub Copilot (a sophisticated software generation tool), demonstrating up to a 55 per cent timesaving on certain programming tasks. We immediately started the rollout of this AI tool to some 3,000 engineers – and our largest take-up has been in the ANZ Plus team delivering our markedly new way of banking.
We have explored a range of business concepts, from augmenting our customer experiences through to helping us with our cybercrime defences. We have seen uncanny improvement in how we can search through documents and policies using an AI pattern called Retrieval Augmented Generation, which allows us to ask natural language questions to find information. We also saw half a dozen groundbreaking concepts demoed at Unconference 2023, our premier internal engineering event.
Finally, we capped out the year with a limited activation of Microsoft 365 Copilot, a desktop AI tool.
What's Coming in 2024
Predicting the future is hard, but the following are for certain:
- Our approach to AI will be deeply steeped in our need for responsibility. Ensuring data security and privacy remains our top priority. AI does not alter in any way our commitment to our values; it reinforces them. AI has the potential to identify bias more effectively than traditional systems, it allows us to make banking safer. The risk of AI abuse by criminal elements is acute, and to protect the banking system's integrity we must harness AI as a tool to counteract these threats. Our Data Ethics principles have been used in ANZ for well over five years and continue to be applied to every single AI use case. We’ve also established a multi-disciplinary AI Risk Advisory forum to review all gen-AI use cases to ensure all concepts meet our risk appetite and values.
- 2024 will be a year where we drive value from AI to advance ANZ’s purpose, aligned to our strategic priorities. Our mission in ANZ is clear, our objectives concrete: we will pursue AI as a mechanism to help us in move more swiftly to achieve our goals.
As I reflect on the past year's journey into the realm of AI, I’m tremendously excited about the year ahead as we apply this technology to our business. I see AI being deployed as a powerful augmentation tool for our staff across all parts of our business. It will enable us to do more with less, freeing our people up to have more time to spend with customers and on the things that make us stronger. We will see this applied to counter cybercrime, to strengthen our internal risk controls, to empower our software engineers to build better technology more quickly. I’m also excited about how we might use this in a way that can enhance our customers’ financial wellbeing.
It’s imperative we – our bank, our industry, our region – are at the forefront of this revolution. Embracing AI responsibly and innovatively is our commitment to a future where technology empowers us to thrive. Together, we are shaping a world where people and communities flourish, and technology is a key force behind that purpose.
Tim Hogarth is Chief Technology Officer at ANZ