Intelligence Vault: Fraud Prevention in Australian Banks

We recently carried out research into the use of artificial intelligence and data analytics within Australia’s major banks — ANZ, CBA, NAB, and Westpac.

This short article in our Distinctive Insights Intelligence Vault series delves into one of the key highlighted themes: Fraud and Financial Crime Prevention.

Our aim in publishing this content is to help finance professionals further understand how artificial intelligence and data analytics are being applied to support key business processes within financial institutions.

More Than Monitoring

In our latest research into the Big Four Australian banks (ANZ, CBA, NAB and Westpac), the theme of fraud and financial crime prevention stood out again and again.

It’s no longer just about monitoring transactions. We’re seeing AI being used to:

• Detect deepfakes and synthetic identities
• Intercept scams before payments go through
• Spot unusual behavioural signals in real time
• Strengthen the link between identity and security

From ANZ’s use of Natural Vision Processing to Westpac’s launch of SaferPay, the shift is clear: fraud prevention is getting smarter, faster, and more human-aware.

Tools, Tactics and Collaborations

Let’s talk about how fraud prevention is actually being done — right now — by the Big Four banks.

It’s not just internal innovation. These banks are actively partnering with AI vendors, building real-time analytics layers, and embedding intelligence into the very edges of their customer journeys.

Some highlights from our research:

ANZ deployed over 170 ML algorithms (already by mid-2023) to flag scam flows, block risky transactions, and spot mule accounts. Partnering with IDVerse, they now verify identities using face biometrics and generative AI — built to detect deepfakes and support fairness.

CBA works Quantium Telstra on tools like Scam Indicator and Fraud Indicator.

NAB, CBA and Westpac are part of a cross-bank behavioural biometrics network via BioCatch — using real-time risk scoring based on device movement and human interaction data.

Westpac launched SaferPay, an AI-driven pre-transaction protection layer. It flags suspicious activity before the money even moves.

What ties these together is a shared mindset shift:

→ From reactive defence to anticipatory intelligence
→ From static rule-sets to adaptive, behavioural models
→ From siloed fraud teams to collaborative ecosystems

Results and Reflections

The Australian banks aren’t just experimenting — they’re scaling these solutions.

From our deep dive, here’s what’s becoming clear:

• AI isn’t just spotting fraud after it happens — it’s reshaping the entire fraud prevention lifecycle
• Behavioural biometrics are becoming as important as payment analytics
• Generative AI is finding its place not just in chatbots, but in ID verification and deepfake defence

And perhaps most importantly: The impact is measurable.

• ANZ saved over $20 million by blocking scam-related flows
• CBA reduced scam losses by 50%
• Westpac processes 30 million interactions daily — with AI models working alongside human intelligence to issue 24/7 fraud alerts

This isn’t just about compliance or cost-saving anymore. It’s about trust, identity, and real-time protection at scale.