Using artificial intelligence to fight fraud and scams

In today’s digital age, particularly fighting fraud and scams, some of the same principles of those ancient safes still apply. They had to evolve over the centuries to keep valuables safe and rebuff security threats. In the same way the modern banking system must evolve and adapt to keep customers and their funds safe.

"Ground-breaking innovations such as AI, machine learning and data analytics are revolutionising our ability to detect and prevent criminal activities.”

According to the Australian Competition and Consumer Commissions (ACCC) Scamwatch, Australians lost a record $3.1 billion to scams in 2022 which is an 80 per cent increase on total losses recorded in 2021. 

Whilst the financial losses for 2023 are trending slightly down on last year, the number of scam reports have increased indicating criminal activity is not slowing down.   

Given the magnitude of the problem it will require collaboration across government, regulators, telcos, law enforcement, social media platforms and the private sector.

At ANZ, we practice defence in depth. Instead of relying solely on processes, people or our technology to protect our customers, we have a number of measures working together to provide the best possible protection, response and resilience to scams and cyber attacks.

No single control alone can prevent an intrusion. It is the sum of multiple controls that makes the difference.

As technology advances, so too does the level of sophistication being used by criminals to target individuals and ANZ. They are innovating faster than ever - using technology to enable mass deployment of hacking, spamming and disinformation. 

Think about your own experience of scams. Those that were once laughably easy to spot with poor grammar, strange email addresses and spelling mistakes, are now harder to detect.

Technology is being used to more accurately mimic individuals and impersonate institutions in attempts to deceive.

Recent media reports have highlighted unethical versions of ChatGPT and deepfake artificial intelligence (AI) that can be used to create images, audio and video. These deep fakes make a targeted attack much more personal, emotive and convincing.

But the rapid advancement of technology has also ushered in a new era of capability in banking.

Ground-breaking innovations such as AI, machine learning and data analytics are revolutionising our ability to detect and prevent criminal activities, ensuring individuals and businesses can transact securely.

ANZ has been using AI and biometric controls such as VoiceID as a second factor authenticator (2FA) in our app since 2018. In the last financial year, we protected $62 billion in transaction value with VoiceID.

We have increased our suite of second-factor controls and deployed additional sophisticated analytical data signals using AI and machine learning, partnering with global leaders who provide best-in-class software solutions.

The sophisticated analytical data signals include behavioural biometrics, malware detection and other device intelligence, bot (robot) detection, remote access takeover signals and mule detection capability in our digital channels.

These passive data signals are very effective in detecting anomalies in a customer’s usual behaviour.

These innovative technology solutions are not ‘set and forget’, they require constant evaluation and refinement of advanced algorithms and rules engines to ensure we have the right balance between security and frictionless digital experiences.  

This means we can strike the right balance between security and seamless digital experiences.