Fraud prevention & security
It is far better for banks to invest now – and deploy advanced AI and ML tools of their own to fight the fraudsters and beat them at their own game.
To mitigate this risk, financial institutions now employ AI systems, in tandem with their human- staffed security operations centres, which are together responsible for monitoring alerts and analysing threats related to fraud, risk and compliance. This approach helps streamline operations and reduce workloads on humans, making them less susceptible to errors, and enhancing overall efficiency. Automated systems, offering capabilities beyond human limitations, act as the initial line of defence – thus minimising the occurrence of errors.
95%
The percentage of cybersecurity breaches that happen due to human error.
IBM 50% Teradata 42
The percentage of true positives detected by Danske Bank’s deep learning AI.
One such financial institution is Danske Bank, which has implemented deep learning and AI to enhance its cyber defences, and ultimately increase the likelihood that it will detect fraud long before it is successful. Recognising that modern-day banking provides a large number of potential attack vectors for fraudulent transactions – including banking on a personal computer, tablet or mobile device – Danske Bank has found that relying on AI and ML has improved its business outcomes, according to a 2018 report prepared by Teradata, a company specialising in AI and ML.
Danske Bank incorporated special software into GPU appliances tailored for deep learning optimisation. This technology helps the analytical model in detecting potential fraud cases, while effectively minimising false alarms. Responsibility for operational decisions is transferred from individual human operatives to AI systems. Nevertheless, human involvement remains essential in specific scenarios. For instance, the model can flag irregularities such as debit card transactions occurring globally, but human analysts are required to assess whether it constitutes fraud – or if the customer is simply on a foreign holiday.
Rise of the machines
By relying on deep learning algorithms, Danske Bank has reportedly seen a 60% reduction in false positives, while also increasing the detection of true positives by half, allowing its human resources to focus more intensely on actual cases of fraud. That’s no mean feat, especially when we consider the fact that fraud costs businesses some $3.5trn annually, according to the Association of Certified Fraud Examiners. A significant reduction in that gargantuan figure means less money in the pockets of criminals – and more finances secured for those legitimately entitled to them. There can be no doubt, at any rate, that AI and ML have transformed the financial industry, ushering in a new era of efficiency and innovation, and there can be no going back. As with any powerful tool, however, they can be misused to cause harm. Financial institutions and banks must strike a delicate balance between harnessing the opportunities presented by these technologies – and mitigating the risks they pose when placed into the hands of bad actors.
As the work of money houses like Danske Bank shows, the march of technological advancement is inexorable. There is no point in wasting time lamenting the use of powerful AI tools by cybercriminals. Rather, the correct approach is to ‘get with the times’ and realise that the technological world is ever-changing and waits for no one. It is far better for banks to invest now, and deploy advanced AI and ML tools of their own, to fight the fraudsters and to beat them at their own game. We may not yet have a benevolent Commander Data from Star Trek to look after our financial interests and wellbeing – but what we do have should still put quite a dent in the ill-gotten gains of cybercriminals. ●
Future Banking /
www.nsbanking.com
metamorworks/
Shutterstock.com
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53
