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. In this piece he looks at whether AI can be used more effectively in the fight against fraud.
Undetected fraud is responsible for generating between £4.6 billion and £10.4 billion of hard losses for the insurance industry every year. Despite the fact that numerous fraud detection measures have been implemented over the years – from operative questioning techniques to case reviews and private investigators – in 2019, the average value of a fraudulent insurance claim came in at £11,400 with an annual industry cost of around £1.3 billion. And with the pandemic creating hard times for all, current data indicates that 2020’s figures could be as much as 21% higher. Reducing these figures is in the interest of everyone, and artificial intelligence (AI) potentially holds at least part of the answer.
How can AI be used to prevent fraud
AI has an enormous remit, but in fraud prevention it is natural language processing (NLP) and automatic speech recognition (ASR) that have the potential to change the game. Because call centres are still seen as the soft target for fraud, with agents being faced with a constant barrage of sophisticated social engineering techniques.
With call centre operatives handling scores of calls a day, each one tasked with processing queries as quickly as possible, the call centre presents an easy touch for even inexperienced fraudsters. Unless a caller has been previously flagged for suspicious activity, or is particularly unskilled at the social engineering requisite to scamming, their fraudulent claim will be processed like any other. Operatives lack the time and training to detect the unscrupulous. But AI has the power to prevent fraud at its source.
Did you know that fraudsters tend to share a common set of speech characteristics? Negation, latency, and emotion are all frequent features of the fraudster’s interaction. They will pause more often, as they think through their answers, trying to avoid inconsistencies. They will react more emotionally if they believe that they are being caught out. And they will negate their sentences. These things sound relatively easy to detect. But if you’re busy, just trying to do your job, they can easily be missed, even by highly trained staff. With conversational analytics, they will be immediately flagged, either directly to the operative, or to the insurer’s fraud prevention team for further investigation. Along with further contextual analysis into the speaker’s credibility, behavioural changes, and emotional state.
Another of the call centre’s unavoidable Achilles heels is the number of people employed. Callers will rarely talk to the same person twice, and this is very much in the fraudster’s favour. The UK Finance IVR’s report stated that a fraudster will make 26 calls to the contact centre during the execution of a given fraud. If the claimant is lucky enough to speak to a different person with each call, the chance of any inconsistencies being picked up is dramatically reduced. With ASR and NPL, every call is monitored, every irregularity noted, and every behavioural pattern identified.
We witnessed this across a piece of work for a UK Insurer where through analysing their fraud calls we were able to show a marked increase in the use of the phonetic alphabet by fraudsters reporting cash for crash claims. The future of fraud prevention
Employed together, NLP and ASR have the potential to stop insurance fraud before it has even started. Of course, they can’t stand alone. They can’t be onboarded and abandoned. Every ASR system needs calibrating for different use cases. And even then, the overall engine can’t make decisions (yet). But together, these systems can provide decision-makers with the real-time, actionable, analytics-led insights that are simply not available from any other source right now. And, excitingly, these technologies have further applications.
Protecting the public against fraud in their own homes has been a long-standing problem. Through collaboration with telecommunications providers, the elderly and vulnerable could be protected from cold callers intent on fraudulent activity, with the application of AI. Because despite widespread publicity, globally telecoms fraud costs around $17 billion a year. In 2018, 25 million scam calls were made to UK residents, according to a BICS report. That’s 15% of the total global telecoms scam communications. With the use of ASR and NLP, these calls could be stopped, defending innocent people from unnecessary loss. And protecting businesses from unnecessary payouts.
Despite enormous recent leaps, AI is still in its early days. In most instances, it cannot replace the human element. And in most cases, nor would we want it to. But with conversational analytics, AI has the potential to provide vital information that could not otherwise be accessed. This has potential in a wide variety of fields. But in insurance, it could prevent billions of pounds of losses, and that has to be worth investigating.