Criminals are exploiting government benefits programs—such as assistance for housing, unemployment or food—to make easy money. The COVID-19 pandemic exacerbated financial pressures faced around the world and government assistance programs were caught in a crossfire of reviewing millions of applications, many that were legitimate and many that were fraudulent. From May to November 2020, the US Department of Justice estimated that $260 million was stolen from 70 cases of Paycheck Protection Program loan scams. But this staggering loss of taxpayer dollars only comprises the cases that were discovered in those six months; there is significantly more fraudulent money to be found from government fraud schemes. And open-source intelligence (OSINT) is the way to find it.
Following the Criminal Digital Footprint
For cyber criminals, online media channels—as well as the deep and dark web—offer anonymous platforms to interact and learn from other bad actors that seek to abuse government assistance programs. Instances of criminal data sharing have risen to such an extent that the U.S. Government Accountability Office estimates that cumulative federal improper payments have totaled approximately $2.2 trillion since fiscal year 2003. Bad actors are aware of the loopholes in these programs, but often these criminals still leave behind breadcrumbs that can be investigated and analyzed.
For many years, fraud was often not discovered until after payments were made to approved applicants; if that money was spent, laundered, or offshored it was typically gone. Over the last 10 years, agencies have implemented more identity verification techniques to validate an identity before the government interacted with it. But as agencies implemented more robust checks and balances, criminals changed their tactics to meet these enhanced fraud prevention solutions.
As scammers adapt and modernize their own efforts, agencies must also modernize and invest in technologies that are able to sift through data at-scale and provide novel insights to ever-changing criminal and cybercriminal behavior. Agencies need to look beyond traditional identity resolution by integrating OSINT solutions into their analytical and investigative toolkits to protect the integrity of their programs and safeguard much-needed money for the public.
Mitigating Government Fraud
OSINT is an integral part of the solution to help identify groups and individuals using the anonymity of the surface, deep and dark web to defraud government benefit programs. Integrating OSINT that is powered by artificial intelligence can aid government agencies in detecting fraud patterns and schemes in a matter of minutes, saving both time and money as manual searches become automated and repeatable.
Systems that leverage AI for web intelligence analysis can bring together large, disparate datasets and networks to detect trends and uncover criminal activity, flagging illicit activity immediately and providing agencies with the information necessary to trace efforts in real-time. These tools are often used to monitor profiles, webpages, Boolean terms, or even multi-signal queries to analyze and cache data in real-time, something that previously required a significant number of analyst manhours. Now, this monitoring can be done in the background while analysts and investigators work on other cases. These AI-powered solutions can search for and detect anomalies that may be indicative of changes in behavior online, signaling potential fraud as well.
Agencies need an OSINT platform that can perform advanced pattern identification, as well as leverage keywords, hashtags, and emojis to provide the requisite granularity into text-based evidence analysis. This type of proactive approach can stop fraud before it occurs, potentially saving the government billions of dollars.
Today’s government analysts and investigators can integrate the latest OSINT into existing analytical and investigative workflows to identify, investigate, and prevent instances of fraud in real-time. Manual identity resolution, laborious activity monitoring, and analysis-by-hand pale in comparison to the outputs generated by automated, AI-powered solutions, both in terms of time required by the analyst and by the potential for uncovering fraud. Using OSINT can help put government aid back into the hands of those who need it the most.
The author, Logan Pauley is an Intelligence Analyst for Cobwebs Technologies.