How speech recognition and AI are fighting fraud in financial services 

Published On : 04/08/2022

By Nigel Cannings, CTO at Intelligent Voice 

Rapid technological advances have created a fast-evolving culture within fraudulent organisations and individuals. Keeping ahead of these fraudulent groups is a challenge for the finance industry, especially as the fast digitisation movement (exacerbated by the Covid-19 pandemic) has led to a significant shift toward the dominance of digital communications. Fraudsters are now able to hide behind a screen or phone, rarely speaking to any employee more than once. 

This makes it far easier for them to fly under the radar – that is unless companies have adopted AI voice recognition technology. The combination of Artificial Intelligence and voice recognition technology is facilitating far more comprehensive anti-fraud cover for financial businesses. It keeps them consistently to speed with evolving fraudulent activity without the need for constant employee retraining and the consistent need for employees to detect fraud themselves (which can detract from good customer service). Overall, speech recognition and AI hold the key to the future of anti-fraud strategy in financial services. 

What roles do machine learning and voice recognition play in fraud prevention?

Combining AI processes with traditional voice recognition technology provides innovative solutions to the constantly evolving nature of fraud in the finance industry. New technology has led to the introduction of the automatic structuring of audio and video data on trade floors and online meetings. The basic processing behind this – GPU-powered transcription combined with NLP (Natural Language Processing) – has been in use since 2014. Businesses are now becoming more aware of how these processes assist in understanding compliance risks (for example compliance with anti-fraud regulations) and sales enablement, all from the same dataset.

Most importantly, by removing the time-consuming and potentially erroneous detection of fraud – and subsequent tracking of behavioural features indicating fraudulent intent – fraudulent behaviour can be identified as early as the first phone call. Machine learning (the process by which computer algorithms evolve independently) allows employees in contact centres to focus their attention on providing the best possible customer service. The potential distraction provided by the need to detect fraudulent intent independently is removed, leaving contact centres fortified with an end-to-end fraud management strategy.

What are the different features recognisable by AI?

AI voice recognition technology such as LexiQal has been developed in cooperation with law enforcement behavioural experts. This ensures the greatest degree of accuracy in automated fraud detection systems. The system itself can detect emotion, tone, speech patterns, specific language features, and more. Features such as negation, hedging, or hesitancy are frequently recognised as key indicators of fraudulent intent, which can be identified and recorded by voice recognition machine learning. Overall, the broad range of features recognisable creates the most comprehensive fraud detection, with minimal involvement from contact centre employees. 

How can voice recognition AI positively impact other areas of financial services?

While fraud detection is understandably a core focus for many financial companies, it is vital to ensure anti-fraud measures do not disrupt the balance with customer service. By automating fraud detection, companies can reduce the scrutiny of genuine customers who may be deterred from using financial services where they feel unduly pressured by employees. Customer service agents will instead be empowered to maintain high levels of service, while fraud prevention measures are present in the background. 

Voice recognition and AI are also becoming increasingly important in maintaining regulatory compliance and reducing fines. Regulatory restrictions are increasingly interconnected, making compliance more complex. This is exacerbated by the post-pandemic boom in the use of other online communications such as Zoom and Teams, which are being examined more closely due to the acceleration of flexible, hybrid working. Applications such as these are featured heavily in both internal meetings and customer communications. Maintaining comprehensive records of these interactions improve a company’s ability to demonstrate compliance, especially when challenged. 

The data collected from voice recognition technology is also applicable to customer protection and sales enablement. Through analysing the success of different approaches used by customer service agents, companies can create a more accurate and measurable sales enablement strategy that places customers at its core. In addition to this, the language and speech detection features of voice recognition AI can be used to identify when vulnerable customers (such as the elderly, adults with disabilities, or young adults) are confused or uncertain. From this information, customer services agents can be made aware of when additional measures need to be implemented to help these vulnerable individuals. More detailed information and explanations of financial policies can be offered, and where necessary, the relevant support services external to the company can be contacted.  

About Author:

Nigel Cannings is the CTO at Intelligent Voice. He has over 25 years’ experience in both Law and Technology, is the founder of Intelligent Voice Ltd and a pioneer in all things voice. Nigel is also a regular speaker at industry events not limited to NVIDIA, IBM, HPE and AI Financial Summits.

Uma Rajagopal

Uma Rajagopal has been managing the posting of content for multiple platforms since 2021, including Global Banking & Finance Review, Asset Digest, Biz Dispatch, Blockchain Tribune, Business Express, Brands Journal, Companies Digest, Economy Standard, Entrepreneur Tribune, Finance Digest, Fintech Herald, Global Islamic Finance Magazine, International Releases, Online World News, Luxury Adviser, Palmbay Herald, Startup Observer, Technology Dispatch, Trading Herald, and Wealth Tribune. Her role ensures that content is published accurately and efficiently across these diverse publications.