Fraudsters vs Algorithms? Tech & Talent Must Combine to Tackle Fraud

In November 2021, Insurance Fraud Bureau (IFB) data revealed that over twenty one thousand fraudulent motor policies were potentially linked to ‘ghost broking’ in the previous 12 months, with the ABI stating almost a six figure sum of fraudulent motor applications being confirmed or suspected.

With the threat of insurance fraud ever-present and with criminals continually probing insurers and brokers to uncover gaps, a mixture of artificial intelligence, data science and digitally skilled investigative colleagues are vital in the industry-wide fight back. Julia Walker-Smith, Associate Director at BGL Insurance, discusses current insurance fraud trends, while sharing her own examples of how technology and talent are being deployed to tackle fraudsters in 2022 and beyond. IFB analysis finds there is an insurance claim linked to ‘crash for cash’ schemes every four minutes – just one of many strategies in the organised fraud playbook, which continually expands and develops.

These strategies often link to topical or current events – indeed, we’re still internally identifying evolving insurance trends caused by pandemic disruption. For example, we saw a direct correlation of insurance fraud spikes every time lockdown restrictions lifted and the teams had to be at the top of their game!

The changing market for fraud: It’s more than controlling premiums, it’s keeping data safe

Digital adoption accelerated during Covid, in particular the use of apps and online portals for self-service and claims reporting. This digital focus increases the threat of account takeovers being used for fraud purposes – such as ID theft, changing the details of a genuine customer, or hijacking a legitimate policy. Also, with vulnerable customers increasingly embracing digital channels, consumer protection becomes even more important.

More established organised fraud tactics also continue to impact the industry. These include ghost broking, where impersonation of legitimate brokers remains a concern, especially given this strategy tends to target more vulnerable communities. As the conflict in Ukraine displaces millions of citizens, there is the possibility of a further increase in the rate of this.

Furthermore, across the industry, issues such as the use of deep fakes for fraud purposes are also emerging – from heavily doctored photos and documents, through to voice re-creation.

Digital to the fore

At BGL Insurance, we already had pre-Covid plans underway to significantly ramp-up and invest in the tech stack in place for counter-fraud operations. This focus has continued, delivering a robust counter-fraud set-up. Under our current system, the models we have work in around 20 milliseconds, with our platform being scalable enough for us to block many fraudulent quotes per second if required.

On the customer side, it’s important to recognise that people still want the convenience offered by digital channels. Therefore, the onus is on us to blend automation with checks for that perfect balancing act of safety and convenience.

By taking this approach, we can save a lot of vulnerable customers from these types of exploitation – without compromising on service levels or customer experience.

Unlocking the full counter-fraud value of data

Although identified motor fraud rates remain low for us – even reducing between 2019 and 2020 while market rates increased – fraudsters are continually evolving their tactics to try and avoid insurer counter-fraud efforts. Data comes to the rescue here, often sourced from price comparison forms, spanning some 140 fields of information for pre-sale fraud analysis. This data alone can be manipulated into thousands of features for fraud identifiers.

But insurers and brokers will first need to achieve data hygiene to unlock full counter-fraud potential. Data must be ethically gathered and analysed in a secure method that respects customer rights and privacy. Firms in the industry will need to enact full data usage reviews to ensure they are harnessing insight for maximum operational value, while ensuring full compliance.

Data can then be used to analyse and understand fraud trends as they emerge, linking these to contributing factors, or topical themes where relevant, to tackle application fraud at the earliest possible stage – and in turn also reducing it at the claims stage.

Don’t neglect the ‘human touch’

In all this, there remains a strong human element required in delivering a successful counter-fraud strategy, which is why BGL Insurance has nurtured more specialist, tech-focused counter-fraud skills and heavily invested in its people – including by recently establishing a new colleague skills and development framework.

We’re seeing demand for training and retaining talent intensifying across many industries where fraud has seen a major uptick. Other industries have found themselves vulnerable to both opportunistic and organised fraud – experienced fraudsters plus new, emerging threats. As a result, counter-fraud skills are heavily in demand across the board.

Algorithms flag fraud at point of quote

Given the sheer volume of quotes and policies we handle every day, there was clear need to introduce advanced data science techniques. AI and machine learning are now a critical part of our counter-fraud strategy.

This focus on data-driven solutions has led to algorithms being deployed to detect fraud at point of quote and granular modelling for post-sale analysis. This has succeeded in cutting detection rates down from an average of 50 days, to less than a week.

But algorithms need to keep up with fraudsters’ tactics. They deteriorate over time – they need daily updates to prevent new fraud tactics slipping through the cracks!

Where next on the fraud front?

We’re certainly aware that in response to new tech and data-driven defences, organised fraud groups are increasing their use of advanced tech such as bots, for example, as a way to apply for large volumes of fraudulent policies.

There will always be an ongoing cycle of fraud trends and so the challenge remains to identify these as early as possible, blocking fake quotes, policies and claims accordingly. Topical scams will likely continue post-Covid, although we expect scam financial services ads to decline now they fall under the remit of the Online Safety Bill, assuming that legislation is passed later in 2022.

Ultimately, as an intermediary, we will continue to prioritise identifying and tackling fraud at the earliest possible stage – all while working to safeguard genuine customers and offer competitive prices, ensuring we remain the go-to distributor for the insurers on our panel.

Advertisement

About alastair walker 9293 Articles
20 years experience as a journalist and magazine editor. I'm your contact for press releases, events, news and commercial opportunities at Insurance-Edge.Net

Be the first to comment

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.