Insurance fraud: Survey reveals impact of economic downturn

The economic downturn seen in the wake of the Covid-19 pandemic has dramatically affected the insurance industry, leading to various challenges in combatting fraud. Worsening conditions have prompted individuals who wouldn’t typically resort to fraud to engage in such activities and provided an opportunity for more sophisticated fraudsters to evolve and refine their strategies.

Cost-of-living fuelling fraudulent activities

According to the survey, an overwhelming majority of respondents (94%) believe the ongoing economic conditions will impact the volumes of organised and opportunistic fraud. As a result, insurance companies face increased pressure.

 

The survey also revealed that opportunistic claims (69.8%) and organised fraud (37.7%) experienced the most substantial rise in fraudulent activities over the past year. These two areas were also seen as the most challenging to identify across the industry, highlighting the need for robust fraud detection measures and prevention strategies.

 

According to those interviewed, a multifaceted approach is needed to tackle these kinds of activities, combining innovative technological solutions with human oversight to tackle fraud on multiple levels.

They claim that by investing in comprehensive fraud training and advanced technology, the industry can ensure the delivery of reliable, cost-effective, and well-rounded services to its customers. This approach enhances fraud prevention and improves insurance operations’ overall quality and efficiency, benefiting insurers and policyholders.

“Preventing increasingly sophisticated fraud typologies can be done by ensuring [the industry] has a range of measures in place,” says Charlotte Brown, counter fraud manager at NFU Mutual.

She explains: “We shouldn’t be solely reliant on technology – sharing broader data with other organisations about individuals is important – but sharing [fraudster] tactics across the industry, overlaid with cutting edge technology, is absolutely crucial.”

We must use a combination of solutions moving forward – it’s no longer a case of one technology fitting all.

Adele Sumner, RSA Insurance

Adele Sumner, head of counter fraud strategy and financial crime at RSA Insurance, emphasised the importance of using diverse fraud technologies to help avoid organised and opportunistic fraudsters.

She states: “We need different fraud solutions and indicators to detect various types of fraud. Improved technology can handle better data and easily process it by quickly identifying trends and patterns.”

Sumner adds: “We must use a combination of solutions moving forward – it’s no longer a case of one technology fitting all. Numerous tools are available, and we need to understand their functions and where they can be integrated into the business to enhance fraud detection.”

Overcoming key fraud challenges

When asked how they measure the success of their fraud processes, referral accuracy and false positive rates were ranked highest by respondents, with false positives seen as one of the biggest challenges in managing their fraud strategy – up 7.6% from 2022.

 

Bobby Gracey, head of fraud at Charles Taylor, believes that false positives are troublesome because they are complicated to catch once mismarked.

“You have to continually refine your front-end selection criteria, the rules, and the weighting around those rules,” he explains. He also notes the importance of “multiple triangulated credentialing factors” to help reduce false positives within current fraud processes.

Greg Cole, claims director of Aioi Nissay Dowa UK, believes that if the technology is built correctly and combined with cross-industry information, there should be a reduction, rather than an increase, in the volume of false positives.

He explains that developing solutions such as machine learning models will help speed up and prioritise the process of false positive detection, assisting insurers to improve their automated fraud detection considerably.

“By bringing intelligence from multiple data sources and combining this with human expertise alongside AI technology, we can prioritise cases most likely to contain fraud,” he says.

“This allows claims handlers to focus on the more important cases, improving automated referral retention overall. It’s about ensuring an effective combination of human versus AI interaction,” Cole adds.

Data sharing: the key to fraud prevention?

Despite the growing volume of fraudulent activities, insurers still encounter challenges accessing broader data, as 58.9% of respondents pointed out (see figure 4). Nearly half of the surveyed individuals expressed their desire or expectation to invest in technology that would provide greater access to shared data within the next 24 months (see figure 5).

 

 

Many insurers struggle to intercept fraudulent activity during key stages of the customer journey. Point of quote was seen as the least successful (41.5%), followed closely by renewal (35.8%), in terms of the data and technology deployed by insurers to address fraud, the survey found.

 

Edward Frost, head of fraud strategy at Axa UK, underlines how financial challenges could contribute to increased fraudulent behaviour from customers making false claims and criminals exploiting vulnerable individuals.

“Advanced analytics and emerging technology such as machine learning can help prevent and respond to these fraud risks,” Frost says.

“Insurers must continue collaborating with industry bodies like the Insurance Fraud Bureau (IFB) to share intelligence and information to fight fraud at every level.

He continues: “Recent innovations in advanced analytical software have helped insurers proactively identify cross-industry patterns and alert the industry to fraudulent activity”.

Yet while data sharing and technology can help prevent fraud across the industry, some insurers still perceive this as a competitive risk.

There needs to be a higher-level response whereby collaboration becomes necessary rather than a ‘nice-to-have’.

Bobby Gracey, Charles Taylor

This raises the question: what should insurers be doing to ensure the progression of broader data sharing across business lines internally and across the industry?

Charles Taylor’s Gracey believes the industry lacks a UK regulator to govern better access to broader shared data across insurance and the wider financial sector.

“There needs to be a higher-level response whereby collaboration becomes necessary rather than a ‘nice-to-have,’” he explains.

“There’s a lot of great work being done by the IFB and other bodies, but for years it has been focused on motor claims.”

Gracey continues: “Fraud is such a vast topic – we need to address it across different lines of business, not just motor.”

However, Cole believes that independent third parties are necessary intermediaries to encourage more data sharing across the industry.

He states: “[Data sharing] should be encouraged wherever possible, and there is an appetite for third-party intermediaries to come in. We must ensure that we maximise opportunities and that [the industry] does not remain complacent.”

RSA’s Sumner points out that the UK insurance industry is already a leader in counter fraud data sharing compared to its EU counterparts: “[The UK] is further ahead than any other insurance market. We have the IFB through which we can share data – yet if you look at other nations, they don’t have the same frameworks in place as we do,” she says.

Sumner continues: “Yes, we can always improve, but I don’t think there is a lack of willingness to share fraud intelligence inside the [UK] insurance market – certainly not among fraud managers in the UK.”

The new frontline for fighting fraud

Although analysing the data gathered through fraud technology helps insurers identify problems throughout the business, it is equally important to have greater firm-wide awareness and engagement when combating fraud.

Fraud solutions can vary significantly across the industry, ensuring that the human aspect of fraud prevention remains integral to the insurer’s fraud processes.

Marie Quinn, head of counter fraud at Woodgate & Clark, emphasises the significance of combining technology with engaged, well-trained staff who can comprehend what they are searching for to guarantee tailored solutions.

She explains how adopting a multi-faceted approach by customising products based on fraud, policy types and client needs is essential for a successful fraud process.

“The industry needs to delve into a more granular level. Once you break it down into this approach, you can start implementing strategic and tactical solutions where data insights become both macro and micro – alongside the human element,” she explains.

The survey identified social media as one of the top three datasets missing from insurer fraud workflows (see again figure 4 above). Some 28.3% (see again figure 5, above) of respondents (an increase of 7.5% from last year) expressed the desire to invest further in social media data analysis in the next 24 months.

The industry needs to delve into a more granular level. Once you break it down into this approach, you can start implementing strategic and tactical solutions where data insights become both macro and micro.

Marie Quinn, Woodgate & Clark

Despite the absence of social media analytics in workflows, the industry recognises its growing role in fraud prevention.

Gracey explains that any sophisticated counter fraud body “…must make intelligence-led decisions” because the intelligence gathered upfront directly affects subsequent investigations.

He provides an example of how insurers can leverage social media to gather intelligence on personal injury claims and identify fraudulent activities, and Charles Taylor recently caught a fraudster using social media data and analytics.

“Social media has a place in fraud prevention. It maximises the potential of analytic technology, but it requires human judgment alongside,” he says, explaining that the risk of false positives significantly increases without human intervention in such technologies.

However, social media plays a vital role in raising awareness about the risks associated with fraud, targeting sophisticated fraudsters and educating vulnerable individuals who may be considering fraud during the current economic climate.

Woodgate & Clark’s Quinn emphasises that insurers must communicate the right message to those contemplating fraud, using social channels to their advantage. “[We] need to encourage the right behaviour – it’s not rocket science,” she says.

Subtle messages like ‘Fraud is the crime everyone pays for’ could help discourage those on the fence about committing fraud.

She continues: “It’s where the education piece comes in, using different types of social media for different target demographics.”

Sumner agrees, stating that the industry needs to be even more proactive in delivering the right message.

“Fraud is not acceptable, and it has consequences. Social media provides a communication channel to make sure that people understand the impact of being involved in fraudulent activity,” she says.

“We have to stay in step with the government and keep communicating about those online harms and how we can protect ourselves and our businesses.”

A new weapon in the fight against fraud?

Generative AI tools such as ChatGPT and Google Bard have become increasingly popular in 2023, prompting new questions about how they will impact the future of fraud in insurance and whether this technology introduces new risks to the fraud space.

Brown highlights that technology like ChatGPT could be employed to generate false posts on social media, thereby creating fabricated histories. She claims it is the industry’s responsibility to “…understand those tactics and ensure everyone is sharing this information [and] aware of these threats and ultimately catch the fraudsters spreading misinformation online.”

“I think it’s a game-changer. People are starting to experiment with it, and there is potential for large language models like ChatGPT to outperform traditional AI and machine learning algorithms,” says Aioi Nissay Dowa UK's Cole.

I think it is too soon to say how AI like ChatGPT will impact us – we can’t predict what strategies will come into play, but we are certainly exploring it.

Adele Sumner, RSA Insurance

He adds, “It can be much more skilled at handling unstructured data and natural language found in claims documents, allowing us to extract more meaningful information and identify patterns and anomalies.”

Despite its potential and growing usage, some believe it is too early to determine the impact of this technology on the industry regarding fraud.

“I think it is too soon to say how AI like ChatGPT will impact us – we can’t predict what strategies will come into play, but we are certainly exploring it,” notes Sumner, adding, “We need to deal with those threats as and when they arise, as eventually fraud detection could turn into AI versus AI, but for now, it is too soon to tell.”

Quinn believes that nothing is off the table regarding fraud prevention and solutions. “There’s no one-stop solution – it’s always about understanding your fraud risk, how it’s committed, and deploying technology appropriate for the type of fraud you are experiencing,” she says.

“A lot needs to be done to develop these solutions. It’s about gathering the right people around the table to ensure your solutions work effectively in specific areas of concern.”

Gracey highlights one of the potential threats of AI. “It can create significant fraud exposures for the industry such as voice mimicry, which is where we need to go back to knowing our customers and asking the right questions. You have to tune in to the behavioural characteristics of the fraudster.

“AI takes [fraud] to a whole new dimension. The industry has been working on utilising AI to try and defeat fraudsters, and I think it’s an interesting concept,” he adds.

In conclusion, the survey revealed that the rise of insurance fraud is a severe challenge for the insurance industry. However, by implementing multifaceted approaches to fraud prevention, insurers can better protect themselves and their customers from this growing problem.

Sponsored