Beyond the Claim: How Social Canvassing is Transforming Insurance Fraud Detection

In today’s digital era, insurers are leveraging social media to gather insights and evidence that could potentially support or jeopardize a person’s claim.

Recently, in the Irish town of Ennis, Kamila Grabska won a Christmas Tree Throwing competition. A photo of her was published in the Irish Independent newspaper, which went viral on X (formerly Twitter), challenging an ongoing personal injury settlement from 2017.

Prior to the photo being published, Grabska asserted that a car accident in 2017 had left her with excruciating pain, rendering her incapable of lifting heavy objects and forcing her to spend days in bed.

She sued an insurer, citing her inability to work and claiming hundreds of thousands of dollars in past and future earnings losses. However, after the photo emerged online showing Grabska making her winning throw, the judge from her case, Carmel Stewart, concluded that the original claim and compensation were inconsistent with the medical evidence presented. “I’m afraid I cannot but conclude the claims were entirely exaggerated,” she said in court.

The details of Grabska’s case are unique, but the outcome is not. Fraudulent claims have long been a challenge for the insurance industry, impacting the trust and integrity of the claims process.

The proliferation of digital footprints and social media has added a new and even more complex dimension to fraud detection and investigation, offering claims professionals new challenges but also valuable tools to uncover deceptive practices.

Behavioral Insights: Understanding and Interpreting Online Interactions and Patterns

Digital footprints encompass a wide range of online activities, including social media posts, geolocation data, digital transactions, and communication records. These digital footprints serve as a treasure trove of potential evidence for claims professionals, offering insights into claimants’ activities, behaviors, and credibility.

In the previously mentioned disability claim case, investigators simply analyzed online posts showing the claimant engaging in physical activities that contradicted their reported limitations. However, information found on social media isn’t always deceptive. Social media and digital channels can also verify claims. For example, a claimant involved in a car accident may post photos or updates about the incident, providing additional context and evidence to support their claim.

People often underestimate the extent of publicly available information. While individuals have control over what they share on social media, they cannot control what others post.

For instance, a friend might share a photo of them on their Facebook page, creating a public record even if their profile is private. This phenomenon extends to family members as well. For example, pictures of one’s volunteer work posted by their spouse or their company’s social media showcasing a team event they participated in all contribute to their online footprint.

However, effective fraud detection requires more than technical skills and common-sense know-how; it also requires a deep understanding of human behavior and psychology.

Fraudsters often leave subtle clues in their digital footprints, such as inconsistencies in timelines, exaggerated claims, or conspicuous gaps in online presence. Claims professionals with expertise in behavioral analysis can use these clues to construct a narrative of potential deceptive behavior, which can strengthen their case against fraudulent claims.

As such, claims professionals must possess a standard set of skills to analyze and interpret digital data effectively, identifying patterns, inconsistencies, and red flags that may indicate fraudulent activity with a full human-level context in mind.

This process involves delving into the layers of social canvassing efforts or systematically gathering information from various online sources to build a comprehensive profile of an individual or situation.

Advanced Strategies: Mining Deep into Digital Layers

To effectively leverage social media in insurance investigations, companies must also invest in advanced analytics tools and train their investigators in digital forensics. But the digital age has transformed social canvassing into a race against a constantly evolving target.

The sheer volume of data across numerous platforms, each with its own quirks, makes staying ahead difficult. Moreover, they must navigate ethical considerations, such as respecting individuals’ privacy rights and ensuring data obtained from social media is used appropriately and lawfully.

To identify actionable leads, claims professionals need to sift through this data, separating valuable information from irrelevant noise.

Locating someone often hinges on understanding their preferred social media haunt. Staying informed about the latest platforms and trends is crucial.

Younger demographics tend towards Instagram and Snapchat, while Facebook skews older. Generalizations aren’t foolproof—some active seniors might surprise you on Instagram or Snapchat, while some people in their twenties may not be on social media at all. Fraudsters constantly adapt their tactics, and yesterday’s go-to platform (think Facebook) might be fading while new ones (like TikTok) emerge. Consider niche platforms too – VK (Russia’s Facebook) for those from Russia or the Baltics, WeChat for Chinese users, or Moj and ShareChat in India .

It is imperative to identify and utilize appropriate platforms to gather evidence that either supports or refutes claims. However, it is equally important to address two crucial concerns: efficient management of vast amounts of video data and the assurance of its authenticity.

While analyzing text and images was relatively straightforward, handling and analyzing hours of video footage on a platform like TikTok or Twitch is a laborious and time-consuming process. Furthermore, with the emergence of deepfakes and other synthetic media types, the line between authentic and fabricated videos has become increasingly blurred.

While AI technology cannot yet fully replace human investigators, AI-powered social canvassing systems can analyze vast amounts of data from social media platforms, public records, and other online sources in real time. These systems can flag suspicious activities, inconsistencies, or discrepancies that may indicate fraudulent behavior. For example, AI algorithms can detect patterns such as unusual spending habits, contradictory information across different platforms, or sudden changes in behavior that raise red flags.

Furthermore, AI can help preserve digital data relevant to claims investigations. By automatically capturing and storing digital evidence such as social media posts, communications, and digital footprints, AI-powered systems create a comprehensive and tamper-resistant record. This preserved data can serve as valuable evidence during claim assessments and fraud investigations, ensuring transparency and accuracy in the process.

The scalability of AI-driven social canvassing enables insurance companies to handle a larger volume of claims while maintaining rigorous fraud detection protocols. Additionally, AI algorithms continuously learn and adapt based on evolving fraud tactics, enhancing their ability to identify sophisticated fraudulent activities.Fraud remains a pervasive challenge in the insurance industry, impacting claims processes and eroding trust. The woman’s win in a tree-throwing competition serves as a stark reminder of the importance of digital footprints and social media in fraud detection. Claims professionals play a critical role in unraveling the layers of social canvassing efforts, leveraging digital evidence and expert analysis to expose fraudulent behavior and uphold the integrity of insurance claims.

Stephenson is director of digital intelligence at INTERTEL, an Ontellus Company.