Maneeza Malik says that if insurers are to become more successful in combating insurance fraud, it is vital that they radically automate and modernise their approach to detection

It is widely acknowledged that fraud costs the insurance industry globally hundreds of billions of pounds each year. According to the ABI, claims fraud in UK general personal lines business alone cost the industry as much as £1bn last year. Unfortunately, most insurance carriers have neither the processes nor technology to stop fraud effectively. At the same time, fraud schemes are becoming more sophisticated and insurance carriers are saddled with an ever-wider array of data to plough through in order to identify fraud trends and detect fraudulent claims.

One way of effectively analysing this mountain of information is for insurance carriers to introduce a set of best practices and techniques that can automatically detect fraud and abuse early on in the claims process. Those insurance companies that already have the ability to develop, deploy and manage applications for automating decisions, find they have a strong basis for a more successful approach to combating fraud. While making a commitment to streamlining the claims process and automating workflows may seem daunting, the benefits of doing so far outweigh inaction.

Detecting fraud
Today, many insurers still rely on inefficient manual processes. Fraud detection is left to adjusters, who may have 30 years' experience or just one, and who have few tools to help them see trends and anomalies despite the high volume of claims. Fraud departments within some carriers use analytics to detect fraud patterns. However, these operations are typically siloed with limited ability to disseminate information throughout the organisation. This provides little assistance to adjusters in their effort to evaluate and take action on a specific claim. For this reason, fraud is often undetected or detected only after a claim has been paid.

Exacerbating the problem is the ever-increasing volume of transactions along with complex multiple reimbursement methodologies that leave insurers wide open to inappropriate claim payments. And there's more bad news. Insurance carriers are under growing pressure to serve consumers through all available channels, including the internet, opening new doors that fraud perpetrators will gladly exploit.

To address these challenges, insurers must deploy a system that provides real-time data analysis on all incoming claims and automates fraud detection processes and procedures. But so far, the insurance industry has been slow to adopt such technologies, in part because existing fragmented, manual processes cannot be easily automated.

For fraud prevention to be effective however, insurance carriers must streamline and automate their claims process and eliminate different claims systems for each line of business.

Using automation
In the fight against fraud, insurance carriers need to be able to automatically assign claims, automate best practices and specific regional regulatory stipulations, as well as determine payment and settlement schemes. Fraud detection can be completely integrated into the claims process and used early on in the claims lifecycle.

Early fraud detection allows insurance carriers to automatically separate claims into 'buckets', detecting anomalies and routing suspicious claims to an investigation group or a suspense file, while letting non-suspicious claims continue through the normal process. This reduces the burden on adjusters to detect fraud, and saves time and money because these claims can be investigated quickly before processing begins.

It is also important that insurance carriers are able to score claims based on multiple attributes and automatically make decisions based on the average score. For example, if a claimant has submitted three claims in the past five years in excess of a specified amount, then the score would be different than if a claimant has submitted a claim below the monetary threshold and has submitted no other claims during the past five years. Alternatively, if a claimant has had a policy for only six months and is submitting a claim in excess of a specified amount, the score could indicate that the claim should be routed to the fraud group.

All of these claims-related 'rules' or decisions can be maintained in a central repository, enabling access across all the lines of business within an organisation. Furthermore, the rules can be changed easily, providing insurers with the flexibility to adapt quickly to evolving conditions as criminals find new ways to commit fraud. Insurers can also take advantage of an improved audit trail, so they can track the 'rules' supporting fraud detection policies over time, which is important for governance and regulatory compliance.

Quality of data
While technology has come a long way in the effort to fight fraud, many tools are still heavily reliant on the volume and quality of the data used in the analysis, detection and prediction. While more data will be acquired as the technologies are adopted, starting with the most accurate and complete set of information will help speed up progress of identifying fraud. Claims provide the greatest source of information on fraud and are the starting place for most fraud detection. Eventually a fraud investigation may lead to fraudulently issued policies, but it remains that the key area for fraud identification is in claims. If there is an opportunity to reduce fraud, it may well lie in the use of technology to identify claims with certain characteristics, as described above.

In reality, no one technology can successfully deliver a complete solution for fraud detection. A complete solution is the result of the intelligent combination of several technologies, most of which are not 100% effective if used in isolation. Those carriers still grappling with manual practices and inefficient processes, which force them into 'reactive mode' in combating fraud, will find however that, with the adoption of technology, they will be able to approach the problem of fraud in a more proactive manner.

Consumers know that insurance fraud leads to higher premiums. Those insurers who are able to cut out fraud, which can be found in as many as one in five claims, will reduce losses and control premium increases and as a result improve their bottom line while creating a clear competitive differentiator.

In other words, if the savings can be passed back to the consumers, then anti-fraud technologies may benefit us all. IT

' Maneeza Malik is industry marketing director of insurance at ILOG