CNseg relies on SAS® to thwart fraud, improves alert accuracy by 67%
Insurance fraud is a growing problem in Brazil. As the Brazilian insurance market approaches R$289 billion (US$70 billion), fraudsters are finding new ways to bilk the system. In 2017, 2.2% of total claims were fraudulent – an increase of about 22.2% over the previous year.
Brazil's insurance confederation, CNseg, is driven to protect more than 100 insurers under its purview. In 2015, CNseg launched a project to centralize fraud prevention operations throughout the Brazilian general insurance sector. The idea was to monitor all claims for suspicious activity and alert insurers of potential fraud so they could take preventative action.
But the process was not very effective. Basically, it relied on the expertise of analysts to manually detect fraud using business rules.
“It was not very organized,” says Ricardo Pereira, Insurance Fraud Prevention and Combat Manager at CNseg. “To truly protect insurers against a swelling army of criminals, we needed a better approach.”
I’m pleased to say that the number of suspicious claims that proved to be fraudulent has increased 67%, thanks to SAS’ fraud-detection solution. Ricardo Pereira Insurance Fraud Prevention and Combat Manager CNseg
Layering on technology
CNseg, which is composed of four major insurance federations in Brazil – FenSeg, FenaPrevi, FenaSaude and FenaCap – turned to SAS for a robust solution to automatically detect both opportunistic and organized claims fraud.
FenSeg – Brazil's national property and casualty insurance federation – is the first federation benefiting from the SAS fraud solution. SAS Detection and Investigation for Insurance uses multiple techniques, including advanced analytics with embedded artificial intelligence (AI) and machine learning, to uncover more suspicious activity than ever before. FenSeg has 20 associated insurance companies that rely on SAS to identify fraud rings that target multiple insurers simultaneously – schemes that each insurer would be unable to detect alone.
Phase one of the implementation immediately took FenSeg’s fraud-detection capability to the next level. By combining business rules, anomaly detection and advanced analytics, FenSeg gained the ability to score insurance claims in real time. Now, when it passes fraud alerts to insurers, investigators have a propensity score to prioritize cases.
FenSeg also gained a visual interface for spotting linked entities and crime rings it may have otherwise missed. “Cluster analysis is a strong weapon in the fight against organized crime,” Pereira says. “It’s much better than the manual method for identifying relationships in the data.”
Machine learning leads to increased accuracy
Phase two of project introduced several new capabilities. Generally, the more data you feed into a fraud scoring engine, the more accurate it becomes.
And perhaps the most powerful new feature was the addition of machine learning. In the past, FenSeg would send fraud alerts to insurers via email and the process to acknowledge if the claim was indeed fraudulent was made through a manual approach. This made it challenging to improve alert accuracy.
SAS changes this by facilitating a two-way connection with insurers via a web portal. Now insurers can input the results of investigations into the system, and a machine learning algorithm will use the data to make the fraud scoring engine more accurate over time.
“Insurance companies can now respond by closing cases, which work as feedback to make the tool more accurate,” Pereira says. “Overall, the system has become more efficient and trustworthy … and we have greater control over the process.”
Since implementing SAS, FenSeg has brought more suspicious claims to light – all through a much quicker, automated and more controlled process. In fact, the volume of alerts has increased by 287%.
“But the key to fraud prevention isn’t just more alerts – it’s more accurate alerts that insurers can act on,” Pereira explains. “And I’m pleased to say that the number of suspicious claims that proved to be fraudulent has increased 67%, thanks to SAS’ fraud-detection solution.”
CNseg – Facts & Figures
Safeguarding the future
Despite its early success, Pereira notes the fight against fraud is never over. Fraudsters are constantly inventing new ways to defraud insurers – and CNseg must stay one step ahead to fulfill its mission of developing and improving the Brazilian insurance market.
“Much still needs to be done,” concludes Pereira. “But I believe SAS is and will always be a relevant partner of the Brazilian insurance market to help us reduce fraud across the sector.”
본 문서에 나오는 결과는 본 문서에 설명된 특정 상황, 비즈니스 모델, 데이터 입력 및 컴퓨팅 환경에 적합하게 되어 있습니다. 각 SAS 고객의 경험은 고유한 것으로, 비즈니스 및 기술적 변수에 따라 달라집니다. 따라서 모든 서술은 비전형적인 것이라는 점을 고려해야 합니다. 실제 절약, 결과 및 성능 특성은 개별 고객의 구성 및 조건에 따라 달라질 수 있습니다. SAS는 모든 고객이 비슷한 결과를 달성할 수 있다고 보증하거나 진술하지 않습니다. SAS 제품과 서비스에 대한 유일한 보증은 해당 제품 및 서비스에 대한 서면 계약의 보증서에 명시되어 있습니다. 본 문서의 어떠한 내용도 추가 보증을 구성하는 것으로 해석될 수 없습니다. 고객은 SAS 소프트웨어의 성공적인 구현에 따라 합의된 계약적 교환 또는 프로젝트 성공 요약의 일환으로 성공 사례를 SAS와 공유했습니다. 브랜드 및 제품 명칭은 각 기업의 상표입니다.