Despite the massive cost of insurance fraud on the industry, many insurers still rely on traditional fraud detection methods that barely scratch the surface of the issue. As a result, insurers only look closely at an average of 4% of all open claims. To help carriers better understand the total landscape of injury fraud and how technology can help close the gap, my team compiled the
The result was a clear indication of when and where fraud can be detected online, and the vital role technology plays for insurers looking to uplevel their fraud detection capabilities.
Where fraud is found online
When relying on traditional fraud detection techniques, adjusters manually scan across internet sources with a trained eye, looking for indications of fraud. However, the reality is that claim-related contradictions can surface across billions of URLs, far more than an investigator could analyze on their own. When AI-powered detection is in the mix, we're able to surface a complete picture of where fraud is found online.
Roughly 17% of fraud is flagged on non-social platforms, such as local news outlets, ecommerce listings, court records, criminal reports, and public databases. However, the highest percentage of fraud (83%) is flagged on social media, consistent with the known societal pressure to overshare to friends, family, and strangers online. This includes popular platforms like Facebook and Instagram, but also non-traditional platforms like TikTok, LinkedIn, and X.
With the largest user base of any social media platform, Facebook was the biggest source of fraud flags in 2024. Facebook is home to a mix of personal updates and location tagging, and as a result, posts about travel, workouts, and lifestyle activities often contradict insurance claims. Video-heavy Instagram and TikTok also surface evidence of physical activity, exposing potential lifestyle and travel-related fraud flags. While not as popular for lifestyle updates, LinkedIn has even evolved into a top source for identifying fraud, as users' posts and updates may unintentionally contradict disability claims or employment status.
Prime times to catch fraud
Surprisingly, the timing of fraud detection proved to be one of the most significant findings in our analysis. Our review revealed a misalignment between when claimants leave digital footprints and when insurers typically begin probing for inconsistencies. This insight reframes how, and more importantly, when, fraud must be addressed.
Of all fraud referrals, 35% occurred in the first two months alone, with the highest alert rate occurring the same month as the injury claim. After the first few months, the frequency of fraud alerts declines, with single-digit percentages each month through the second year. Yet, while the alerts might taper off gradually, they never hit zero. Almost 50% of all fraud is still found as late as two years after the first notice of loss. Typically, this activity is related to life changes, such as new jobs, starting a business, or sharing updates that contradict the original claim.
What carriers should know
The spike in detection in the first few months validated two main findings. First, it proved that the traditional fraud detection processes were too late to the game; claimants often make their most revealing online updates before adjusters, SIU, or counsel are brought in. Second, the data demonstrated why investigations can't be a one-time thing. By relying on a single point-in-time social media check or waiting for something to "feel suspicious," the moment when evidence is most visible will have passed.
Ultimately, the findings showed the importance of proactive and continuous assessment of online content. The landscape around fraud is changing, and online evidence needs to be reviewed with a broad and lasting lens. And this isn't something insurance professionals can perform efficiently on their own; they need tools in hand that provide consistent, compliant, real-time visibility from the moment of loss through the full lifecycle of the claim.
