Fraudsters share intelligence every day. Why don’t insurers?

Fraudsters have become more connected, more technologically enabled and more sophisticated. The insurance industry knows it needs to respond. Today’s criminal networks can share information quickly, exploit weaknesses across multiple insurers and adapt quickly to new controls. Insurers meanwhile are still struggling to turn the intelligence sitting across the market into a collective defence.

This is one of the clearest messages emerging from research by CRIF and Insurance Post. The problem is not a lack of technology. Nor is it a lack of willingness to collaborate. The problem is connecting intelligence quickly enough to make a difference.

The nature of insurance fraud is changing

According to the research, 68% of respondents believe fraud has become more technologically enabled over the past three years. More than half, notably claims professionals, say it has become more opportunistic and volume driven. Meanwhile, fraud professionals are seeing the growth of organised and technology-enabled activity.

Both perspectives carry equal weight because whether the threat is an inflated claim from an individual policyholder or activity involving a coordinated fraud network, insurers increasingly need to identify connections between pieces of information that may look insignificant on their own.

AI isn’t enough

It would be easy to conclude that the answer is simply more sophisticated technology. The industry is already moving in that direction. More than half of respondents are using AI-driven tools to detect and prevent fraud, while 87% believe AI-driven detection will be one of the most important developments over the next five years. But AI does not operate in a vacuum.

Its effectiveness depends on the quality, breadth and timeliness of the intelligence available to it, in addition to human oversight. An insurer can have sophisticated internal analytics, but if the information needed to identify a wider pattern sits elsewhere in the insurance ecosystem, technology alone cannot close the gap.

The industry wants to collaborate

Fraud does not respect organisational boundaries. An identity, business, address, account or pattern of behaviour can appear across multiple interactions within the insurance market. The more fragmented the industry’s view, the easier it is for those connections to remain hidden and for fraudsters to exploit the gaps. Almost nine in ten respondents say collaboration is either critical or very important to tackling fraud effectively.

Fraud does not respect organisational boundaries. An identity, business, address, account or pattern of behaviour can appear across multiple interactions within the insurance market.

The research reveals an important insight into how collaboration currently takes place. Most external fraud data is shared during active investigations. Much less sharing takes place at the pre-claim or underwriting stage. Insurers are typically sharing intelligence once they already suspect something is wrong, rather than using collective intelligence to prevent questionable business entering their book in the first place.

The earlier relevant intelligence can influence a decision, the greater its potential value. Going forward there is opportunity for it to become part of the decision-making infrastructure that helps identify risk earlier. That means connecting internal data with trusted external intelligence, embracing a standardised and compliant digital platform for sharing intelligence, using AI to identify patterns and making relevant information seamlessly available at underwriting, FNOL and claims stages.

Process friction

The industry already has established mechanisms for sharing information, including centralised databases, formal data request processes and third-party data and analytics providers. The effectiveness of these mechanisms can be varied, however.

Formal data requests are unquestionably challenging. A majority of respondents use structured formal data requests either frequently or sometimes. But the process is fraught with friction. The most common frustrations are delays in receiving responses, the lack of a streamlined digital process and inconsistent responses between organisations.

For a fraud professional investigating increasingly sophisticated activity, waiting for information or receiving inconsistent responses can mean that valuable intelligence arrives too late to influence a decision.

The research suggests the industry understands this. Some 62% strongly support the idea of a secure, standardised digital platform for sharing intelligence, with a further 28% somewhat supporting it. The appetite for a more connected model is considerable.

Better collaboration benefits more than fraud prevention

Survey respondents did not see better data sharing simply as a way of catching more fraud. They expect stronger collaboration to improve fraud detection rates and reduce losses, but also to deliver faster claims resolution, lower operational costs and better customer outcomes.

Fraud prevention and customer experience are sometimes treated as competing priorities. Better intelligence suggests they can actually reinforce each other.

The ability to identify higher-risk activity earlier means resources can be concentrated where they are needed, while legitimate customers can benefit from faster and more confident decision-making.

The next five years will be about connection

The research points towards a clear future model. AI plus improved intelligence plus collaboration. AI will undoubtedly become more important. But the insurers that gain the greatest advantage will be those that can combine AI with richer, reliable external intelligence and effective collaboration across the ecosystem.

The insurance industry recognises the threat from connected criminal networks. It already has much of the technology and the appetite to work together. The challenge now is turning that willingness into a genuinely connected intelligence defence.

Sponsored content