AI-Native Fraud Detection

AI-Native Fraud Detection for the Global Insurance Industry THE SITUATION: INSURANCE FRAUD COSTS BILLIONS GLOBALLY 10-20% of all insurance claims contain some element of fraud, but only a small fraction of that fraud is detected. Globally, insurers stand to lose hundreds of billions in revenue to this pervasive issue. Historically, detecting fraud has been a costly, inefficient process. That’s why we built FORCE. It’s the only AI-native, SaaS-based fraud detection solution in the world, trusted globally by leading insurers to detect fraudulent claims and organized fraud networks accurately and efficiently at scale. AI + SAAS + WORLD CLASS DATA SCIENTISTS = FORCE FORCE HIGHLIGHTS FORCE detects and provides context for a wide range of fraudulent activities, from opportunistic claim exaggerations to schemes executed by organized crime networks. It’s a SaaS+ solution that combines sophisticated AI and human data science expertise, HIT RATE replicating and multiplying the deductive capabilities of an insurer’s best fraud handlers. 35% (Industry SUSPICIOUS average) CASES AI-NATIVE FRAUD DETECTION WITH A 75% HIT RATE Vs Once FORCE is configured, each claim is run against an evolving library of hundreds of HIT RATE 75% fraud scenarios – statistical patterns trained on historical datasets at the customer and Shift industry levels – to detect claims matching existing fraud patterns. FORCE then applies additional AI techniques to each claim such as “privileged learning” algorithms and natural HIT RATE = language processing (NLP) of text variables. N° OF ALERTS DECLARED SUSPICIOUS BY THE FRAUD HANDLER rd Next, FORCE uses a wide range of 3 party data to develop a full picture of each claim N° OF ALERTS SUBMITTED BY FORCE with speed and accuracy at scale. From weather conditions and satellite imagery to social media monitoring and more, FORCE contextualizes and expands claim data. It then uses this enhanced body of data to assign each claim a numerical score representing • AI-native SaaS solution the likelihood of fraud. • Built by world-class Data Scientists The solution delivers automated alerts to internal team members, who can access claim with deep industry expertise data through FORCE’s flexible, intuitive UI. In addition to scoring, FORCE flags specific • Hundreds of fraud scenarios aspects of the claim for further review, supplies reasoning for all fraud indicators in clear constantly evolving business language, prioritizes investigative avenues, and presents the ‘packaged’ individual claim and network fraud data in a single, customizable dashboard. • Hundreds of millions of claims processed to date FEWER FRAUDULENT CLAIM PAYOUTS, INCREASED EFFICIENCY • Addresses individual claims + FORCE helps insurers detect and reduce payouts on non-meritorious claims accurately network fraud and efficiently. It speeds claim resolution and keeps customer teams focused on the • Transparent Analytics access activities where their skills will be most valuable. At the same time, FORCE playes an to scenarios + data important role in straight through processing (STP), enabling Shift’s Luke solution to fully • ROI in as few as four months automate claims without increasing the risk of fraudulent claim payments. FORCE also identifies and maps organized fraud networks, providing detailed, actionable information to address this critical problem. shift-technology.com

AI-Native Fraud Detection for the Global Insurance Industry 2 THE FORCE SAAS+ MODEL: KEY ADVANTAGES FAST, EASY ONBOARDING To get started, insurers simply provide historical policy and claims data in any format. FORCE ingests, cleans, and maps the data. Shift Data Scientists manage the process, working closely with clients' internal teams to fine tune the solution before proceeding to full deployment in as few as four months. LEVERAGES THE POWER OF YOUR INTERNAL DATA FORCE’s AI-native foundation is built on insurers’ internal data, including claims data, policy information, financial data, loss adjuster reports, damage estimates, medical billing data, photos, and more. BENEFITS FROM YEARS OF DEVELOPMENT OF FRAUD SCENARIOS FORCE has been in continual development and refinement since 2014. In that time period, the solution has incorporated massive amounts of existing claims data – from hundreds of millions of claims – which form the basis for more than 200 highly-defined fraud scenarios across multiple insurance categories. ENRICHES SCENARIOS WITH ROBUST THIRD-PARTY DATA FORCE ingests and analyzes information from sources such as weather data, photos, satellite imagery, criminal records data, bankruptcies, liens, judgements, IP addresses, social media data, loss history, and even data derived from automated web crawling of publicly available websites. This supplements the model scoring by creating and weighting new variables, providing unmatched insight into potential fraud. GATHERS AND INTERPRETS UNSTRUCTURED DATA FORCE analyzes unstructured data (such as claim adjuster notes) using text mining and natural language processing to identify keywords and phrases that are indicative of fraud, leveraging this text analysis to create fraud variables that will be used in scoring the claim. PROVIDES THE EXPERTISE OF WORLD-CLASS DATA SCIENTISTS Shift’s team of more than 100 Data Scientists collaborates with clients’ internal business and data science teams to refine a joint approach to fraud detection throughout the entire customer lifecycle. Our transparent approach to the sharing of data (and the variables and scenarios that comprise the detection model) means that customers’ internal data science teams can develop their own rules, scenarios, and features to incorporate into their FORCE configuration. DELIVERS ACTIONABLE CLAIM ALERTS VIA AN INTUITIVE, CUSTOMIZABLE USER INTERFACE FORCE provides prioritized, scored individual and network fraud alerts to claim handlers, fraud handlers, and SIU investigators. They can easily view and manage fraud scenarios and indicators, along with third-party data, maps, and network visualizations in a single interface, without having to refer to multiple data sources externally. Investigators can quickly determine the validity of claims and prioritize their efforts where the ROI is greatest instead of wasting time on triage and research. AUTO HEALTH HOMEOWNERS’ TRAVEL WORKERS' DISABILITY LIFE COMPENSATION Tap into the power of AI-native fraud detection today. Visit shift-technology.com shift-technology.com