As fraudsters become more sophisticated, Steve Paton explains how insurance companies can use data analytics and technology to stay one step ahead
As insurers move towards automation to meet today's digital-first consumer expectations, there is much talk of delivering a 'seamless customer experience' and 'shortening the claims life-cycle'. However, legacy systems can struggle to keep pace with these customer-centric goals. Fraudsters have been quick to adapt, and new types of fraud have started to emerge, exploiting insurers' desire to reach fast settlement.
Fraudulent claim attempts detected by insurers have now reached an annual value of £1.28bn in the UK, and no one expects fraudulent activity to slow down in this fast-changing market. Some commentators are already predicting that fraud cases could double in the next few years, as some insurers opt to lower claims payment thresholds to achieve faster, low-touch settlements.
If insurers are to keep pace with changing fraud schemes, they need more than ever before to to be able to adapt quickly. They must have an innovative approach to fraud detection - one that incorporates emerging technologies, agile solutions, and real-time shared intelligence across claims tools for better analysis and interrogation.
Evolving fraud patterns
Recent legislation in the UK, such as the Civil Liability Bill, which received Royal Assent last December, could complicate the challenge of balancing fast settlement of meritorious claims with the due diligence of fraud detection. Among other reforms, the law establishes new whiplash measures that raise the small claims limit to £5,000.
These types of claims are considered low risk, and insurers often fast-track them. However, raising the thresholds on these claims elevates the risk of fraudsters gaming the system - especially organised criminal gangs, which can file several small claims through a large network of conspirators. Eventually, those small settlements add up and become costly for an insurer.
Despite the improvement of counter-fraud measures, fraud schemes persist and can be a huge drain on an insurance portfolio's profitability. Known tactics, such as contrived accidents, still inflict a heavy cost on insurers.
Long-established schemes involving professional enablers, such as doctors, solicitors and engineers, also persist. Fraudsters routinely abuse their positions and make a sustained effort to disguise the fraudulent nature of a claim and then progress it.
Then there's the insider threat. Insurance employees with an intimate knowledge of counter-fraud systems and their thresholds can facilitate fraudulent claims submissions that may go undetected. Addressing these challenges requires sophisticated counter-fraud solutions.
Anti-fraud capability gaps
Advanced data analytics can identify many fraud schemes in the early stages of a claim, and most insurers employ some type of counter-fraud system. However, whether an insurer uses an internal system or a vendor solution, there are inherent limitations with many current tools.
One common vulnerability lies in legacy systems that may not be fully integrated into all other related systems, creating silos of valuable data. Such systems are especially vulnerable when fraud patterns change. Shrewd fraudsters can quickly identify and exploit loopholes in new regulations because they have a deep understanding of the counter-fraud measures and broader claims processes currently in place.
To counteract the increasing complexity of fraudulent claims and system challenges, newer counter-fraud solutions use advanced data analytics, configurable rules-based identification, machine learning and document scanning to detect fraud patterns. They draw on extensive datasets to discover the slightest anomalies, share intelligence between systems and, subsequently, flag claims for further review. Insurers with minimal IT capacity, limited budgets and siloed data could use cloud computing technology to deploy these advanced capabilities within a short time-frame.
It's important to note that these solutions don't supersede human expertise - rather, they support 'right touch' claim handling. The technology focuses on delivering advanced capabilities to supplement the knowledge of fraud investigators and claims handlers. By identifying complex patterns, it aids the decision-making process and boosts operational efficiency.
Effective counter-fraud tech
A core requirement of a counter-fraud technology solution is the ability to establish a secure centralised repository - with hashed data to ensure security - that will process, store and distribute the data in a format that is fully compliant with regulations and best practices, such as GDPR and the National Intelligence Model. This enables insurers to closely collaborate on intelligence sharing and fraud detection, which will ultimately help increase detection rates and the efficiency of investigations.
Beyond data handling, a counter-fraud system must operate in real-time, instead of simply being deployed for specific checks on suspicious claims. This involves continually scanning and assessing policies for fraudulent patterns of activity - whether this is at the time of quotation, policy inception, midterm amendment, claim or renewal. This system will also ideally link directly into existing policy, claims and other third-party systems, such as those operated by preferred suppliers. This will, for example, enable insurers to incorporate an automated check for valid MOT and check mileage to see if it differs from the claimant's report.
Suspected fraud must be accurately flagged for investigation. This means the most advanced systems apply visual link-analysis technology to standardised, highly searchable data to uncover previously undetected suspicious patterns and refer them to fraud investigators. These patterns and associated data can also support broader anti-fraud partnerships by being shared between insurers, law enforcement agencies and other industry organisations.
Looking beyond claims and into the applications for wider business operations, an effective counter-fraud system also helps insurers identify and avoid insider threats as early as the recruitment and onboarding stages, by cross-referencing details such as addresses and phone numbers with those associated with known fraudsters.
Minimising business disruption
Insurers that integrate advanced technology into claims processes will be best placed to cope with these ongoing industry challenges - and react more quickly to mitigate any potential impact to their bottom lines.
While new, innovative solutions are intriguing, the insurer must consider the extent to which a large-scale digital transformation, overhauling existing claims systems, could cause business disruption. This risk is significantly mitigated through the selection of the right counter-fraud technologies - those that have been developed with agility and smoother integrations in mind.
Cloud-based solutions go a long way to help minimise business disruption with easy implementation. The small IT footprint of these solutions also brings lower maintenance costs, helping insurers' bottom line.
As the insurance digital landscape continues to evolve, fraud detection strategies need to stay ahead of the curve. Embracing the very latest agile technology and real-time intelligence sharing and analysis is now essential across the policy life-cycle.
Steve Paton is head of Anti-Fraud Services (EU), Verisk Claims