Prevent fraud, without friction.
The U.S. Government Accountability Office estimates that as much as $521 billion is lost to fraud annually. Fraudsters use advanced techniques like bots, stolen and synthetic identities, IP spoofing software, and the dark web to conceal fake applications, scale operations, and traffic stolen identities. In 2023 alone, over 353 million Americans were affected by data compromises. That’s on top of the 5.5 billion compromised records over the previous five years. This means that the average American’s records have been exposed 18 times since 2017, creating a target-rich environment for sophisticated fraud operations.
Given this environment, governments should consider the following actions to improve program safeguards. These will help fight fraud, reduce backlogs, expedite processing, and protect privacy and Personally Identifiable Information.
Focus on pre-payment prevention versus post-payment detection.
While post-payment “pay and chase” fraud activities are necessary for program integrity, they are expensive and mostly ineffective. They often collect as little as 5% of total fraudulent payments. Pre-payment solutions effectively “collect” 100% of the fraud they identify. The only meaningful expenditure is the cost of the prevention software.
Automated pre-payment technologies should be included as the first line of defense.
Pre-payment technologies can effectively prevent most fraud. This includes large attacks that use stolen and synthetic identities, drop houses, bots, and other fraud-scaling techniques.
Use behavioral analytics to improve identity verification capabilities.
Organized fraud rings depend on one universal capability: the ability to access, submit, and conceal stolen and synthetic identities. This includes mimicking demographic information, disguising geographic locations, and even information about the device they use to complete their applications. Unfortunately, bad actors have become very good at using IP spoofing software, headless browsers, and other technologies to cover their tracks. Fortunately, behavioral analytics can identify bots and stolen identities by analyzing the speed, patterns, navigation, and typing techniques used to fill out an application. Each visit to your website generates thousands of unique data points, whether someone is applying for a benefit, re-certifying, or modifying their account information. This data is invaluable for confirming identities. Behavioral analytics offers a comprehensive identity validation solution when combined with demographic and device data. It provides a digital fingerprint to differentiate between a legitimate applicant and a fraudster. Put another way, it’s a lot easier to copy someone’s SSN than it is to copy their fingerprint.
Reduce false positives with comprehensive triangulated fraud indicators.
Rather than relying on “single indicators” of potential fraud, such as public records checks, it’s important to form a more comprehensive view of an applicant. This includes not just their demographic information but also their device data and their behaviors. This data can be combined and scored using modern technologies to return more accurate and actionable applicant risk scores.
Improve program integrity while protecting citizen privacy and streamlining access to services.
One of the most important lessons of the pandemic is that governments can eliminate fraud, but it would come at the cost of privacy and service accessibility. For example, requiring in-person application interviews, facial recognition checks, or even physical fingerprinting goes a long way toward stopping fraud. But at what cost? As these and similar techniques were adopted during the pandemic, advocacy groups filed lawsuits claiming privacy violations, unequal access to services, and bias in models used for facial recognition. New technologies like behavioral analytics and advanced hashing algorithms allow you to fight fraud without sharing PII with outside vendors, slowing processing times and subjecting constituents to invasive processes and burdensome identity checks. These more defensible technologies allow you to protect program funds while also protecting constituent privacy.
Summary
No single fraud detection process or technology will eliminate fraud in government programs. Fraudsters encouraged and emboldened by their successes before and during the pandemic, continue to embrace cutting-edge technologies to target lucrative subsidy programs. On the other hand, government agencies continue to operate from the difficult position of quickly processing applicants, minimizing false positives, protecting program funds, and protecting citizen privacy and data. One of the most effective ways for governments to accomplish these goals is to use behavioral data to identify fraudulent applicants before they enter the programs or the agency’s workflow.