Making AI fit for purpose

The federal government aims to make Australia a global leader in responsible and inclusive artificial intelligence (AI). A key tenet of the sound use of AI is ensuring it is ethical. To help businesses meet this requirement, the government has created a voluntary AI Ethics Framework to guide responsible design, development and implementation.

Responsible application of AI is something the Bank is committed to as it continues to enhance its technical capabilities. To that end, we recently participated in a government-led pilot program to test its ethical AI framework, along with several other large Australian businesses.

The framework encompasses eight key principles:

1. Human, societal and environmental wellbeing: AI systems should benefit individuals, society and the environment
2. Human-centred values: AI systems should respect human rights, diversity, and the autonomy of individuals
3. Fairness: AI systems should be inclusive and accessible, and should not involve or result in unfair discrimination against individuals, communities or groups
4. Privacy protection and security: AI systems should respect and uphold privacy rights and data protection, and ensure the security of data
5. Reliability and safety: AI systems should reliably operate in accordance with their intended purpose
6. Transparency and explainability: There should be transparency and responsible disclosure so people can understand when they are being significantly impacted by AI, and can find out when an AI system is engaging with them
7. Contestability: When an AI system significantly impacts a person, community, group or environment, there should be a timely process to allow people to challenge the use or outcomes of the AI system
8. Accountability: People responsible for the different phases of the AI system lifecycle should be identifiable and accountable for the outcomes of the AI systems, and human oversight of AI systems should be enabled.¹

CommBank Chief Decision Scientist, Dan Jermyn, says the government’s principles are important because they specifically call out concepts such as fairness, transparency and explainability, which can be overlooked when developing complex AI systems.

“Our organisation is committed to using technology, data and AI safely and with the right controls in place to help improve our customers’ financial wellbeing,” Jermyn said.

CommBank uses AI to power its Customer Engagement Engine: a system that analyses data in order to personalise the way that the bank manages its communication with customers, and learns from every interaction. 

That learning helps to continually improve the customer experience, and provide enhanced products and services. The constant testing and experimentation that happens in the background is just one of the ways we ensure our application of AI adheres to best practice.

“Our well-established processes and governance, including around data safety and security, help CommBank to apply AI safely and with the right accountabilities,” said Jermyn. “[The pilot program] was a good opportunity to share general learnings and experiences about AI, as well as identify opportunities for potential industry improvements.”

The government said pilot participants all agreed that pressure for AI to be ethically designed and deployed will grow, fuelled by public pressure and government or regulator policies.

Other key learnings from the pilot program included:²

• That the responsibilities of AI purchasers differed from AI developers, i.e. businesses buying AI solutions recognised they could not outsource their accountability for AI ethics. To meet those principles they relied on internal due diligence and information from their vendors –information on how the model makes decisions; on the data it has been trained on to minimise risk of bias; and to understand the limitations of the model across a range of applications.
• Developers of AI found it challenging to manage the ethical impacts that occur after they sell their AI system and no longer control its deployment.
• Where systems are not appropriate for the application intended by the purchaser, both parties (developer and purchaser) should consider getting the AI system re-designed.
• With regard to staff, businesses need to raise awareness of AI ethics and educate about the benefits of implementing AI ethics for the company and its customers.
• Improve governance processes – include AI applications in risk assessment processes and data governance.
• Businesses also found that some principles were more challenging to practically implement. Principles such as fairness, for instance, can be harder to judge and measure.

More information on CommBank’s participation in the pilot program can be found here.