Congress provided more than $2.3 trillion in new funding through three bills: The Infrastructure Investment and Jobs Act (IIJA), the Inflation Reduction Act (IRA) and the Creating Helpful Incentives to Produce Semiconductors Act (CHIPS and Science Act of 2022). Grants have been the preferred procurement vehicle for many federal programs, as they provide increased flexibility to recipients, help advance public purposes, only require “best effort” (vs. a specific delivery) and are not performance-based. And guaranteed loan programs were the target of historic fraud during the pandemic, highlighting the need for oversight. However, grant and loan programs have been woefully lacking in fraud prevention resources to date, thus raising the risk that these new programs will be rife with fraud.
The IRA authorizes about $250 billion in new loan guarantees and increased current loan guarantees by $100 billion. It also provides $3 billion for direct loans to finance low-emissions vehicle manufacturing, increases the authorized capacity for loans to finance energy projects for Indian tribes, establishes a new program to finance energy infrastructure projects and allocates funds for credit subsidy costs of direct loans to build new transmission lines and modifies existing ones.
The CHIPS and Science Act provides more than $52 billion for American semiconductor research, development and production.
The IIJA provides $550 billion in new federal investment in roads and bridges, water infrastructure, resilience and internet access. And much of this funding will flow through grant and loan programs.
While the IRA introduces a new layer of oversight by requiring the President’s certification for loan guarantees and compliance with relevant provisions — including a prohibition on double dipping — none of the acts include any noteworthy provisions to prevent fraud.
Playing the data game
New programs in new industries are easy targets for fraudsters, and environmental initiatives — like building renewable energy infrastructure — are still very new in many parts of the country. The Environmental Protection Agency’s enforcement division has brought 16 renewable fuel fraud cases in the last 10 years, with many more cases referred to the Justice Department for criminal prosecution. In 2019, members of a religious sect known as “The Order” pleaded guilty to conspiring with a Los Angeles businessman steal $1 billion in a scheme involving renewable fuel standard credits and related IRS tax credits that involved multiple shell companies. These schemes are about to multiply.
Consider this scenario: An enterprising fraudster decides to create five new shell companies purportedly aimed at building renewable infrastructure. He enlists four friends to put the other companies in their names, so it looks like a competitive group of companies applying for grant funds. They create bogus past performance documents attesting to their qualifications and set to work applying for grant funds in 23 states through the Department of Energy’s Energy Efficiency and Conservation Block Grant, which is now receiving $550 million in funding through the state and community energy program. Despite receiving an influx of billions in new funding, oversight for these programs remains significantly under-resourced.
Overworked and under time pressure, the adjudicators check to be sure all the required documentation is included. The ring uses the same past performance information on all the grant applications, which would be a red flag, but only if the agency had tools to compare text in grant applications across states. Maybe the adjudicators run the companies through a cursory check to ensure they haven’t had any fines or criminal charges levied against them, and since the companies are all new, there are no concerning records that would flag them as suspicious. Because the owners are all different, and the adjudicators don’t have any tools to identify the relationships between the business owners, they have no reason to suspect a ring is involved. With the documents all in order, the adjudicator awards the grants. Once the initial grant payments are deposited into the fraudster’s bank accounts, he closes the accounts and disappears.
AI/ML are key technologies
Preventing fraud is a data game. Bad actors practice “eligibility laundering” across grant programs using rings with small businesses, non-profits and foreign entities all trying to conceal, through deception, their unqualified status within the programs they’ve applied to. Combating these fraudsters requires data, and the tools to use that data quickly and effectively.
The acceleration of machine learning and artificial intelligence tools offers government agencies the ability to identify frauds like the scenario above quickly. Natural language processing text analytics engines can identify duplicate passages in grant applications in seconds. Massive amounts of third-party data can be mined and leveraged to identify past criminal activity and other suspicious indicators related to applicants. Third-party data analysis can also identify patterns indicative of stolen or synthetic identities used in grant and loan applications. Social network analytics can identify the relationships between grant and loan applicants that could indicate the existence of a fraud ring. Device metadata, such as geolocation, can also be mined to identify potential fraud indicators. And all this data analysis can be done in seconds with the tools available today.
If we learned one thing from the pandemic it must be this: Fraud actors have the upper hand. Pandemic relief programs were fleeced in large part due to the lack of data-driven tools and capabilities across federal and state agencies. To protect taxpayer resources, government agencies must invest in the tools needed to fight fraud. For those bills already funded and being implemented, agencies should evaluate their internal capacity to conduct some of the analysis described above, and consider options for redistributing funding to prioritize investment in fraud prevention tools. Going forward, Congress should establish funding for program integrity to accompany large spending bills and direct agencies to establish metrics for preventing fraud actors from stealing the funds. It is simply irresponsible to continue dedicating hundreds of billions of dollars to new spending programs without dedicating the investment in oversight tools to protect the integrity of those funds.
Linda Miller is the founder and CEO of Audient Group, LLC and a nationally recognized expert in fraud risk management. She was a partner at Grant Thornton, LLC where she led the firm’s fraud and financial crimes practice and formerly served as the deputy executive director of the Pandemic Response Accountability Committee (PRAC). She also led the development of the Government Accountability Office (GAO) Framework for Managing Fraud Risks in Federal Programs.
Erik Halvorson is the owner of Archimedes Evaluation Services, LLC, where he supports non-profits and businesses in analytics, program evaluations, and anti-fraud program management. He has 18 years of federal law enforcement experience, including serving as a special agent.
More federal spending should come with investment in data-driven tools to fight fraud
Linda Miller and Erik Halvorson, fraud prevention experts, explain how new technology tools and data can help protect agencies from fraudsters.
Congress provided more than $2.3 trillion in new funding through three bills: The Infrastructure Investment and Jobs Act (IIJA), the Inflation Reduction Act (IRA) and the Creating Helpful Incentives to Produce Semiconductors Act (CHIPS and Science Act of 2022). Grants have been the preferred procurement vehicle for many federal programs, as they provide increased flexibility to recipients, help advance public purposes, only require “best effort” (vs. a specific delivery) and are not performance-based. And guaranteed loan programs were the target of historic fraud during the pandemic, highlighting the need for oversight. However, grant and loan programs have been woefully lacking in fraud prevention resources to date, thus raising the risk that these new programs will be rife with fraud.
The IRA authorizes about $250 billion in new loan guarantees and increased current loan guarantees by $100 billion. It also provides $3 billion for direct loans to finance low-emissions vehicle manufacturing, increases the authorized capacity for loans to finance energy projects for Indian tribes, establishes a new program to finance energy infrastructure projects and allocates funds for credit subsidy costs of direct loans to build new transmission lines and modifies existing ones.
The CHIPS and Science Act provides more than $52 billion for American semiconductor research, development and production.
The IIJA provides $550 billion in new federal investment in roads and bridges, water infrastructure, resilience and internet access. And much of this funding will flow through grant and loan programs.
Join us May 6 at 1 p.m. EST for Federal News Network's Industry Exchange Data where we'll explore how you can employ data to modernize your agency. | Register today!
While the IRA introduces a new layer of oversight by requiring the President’s certification for loan guarantees and compliance with relevant provisions — including a prohibition on double dipping — none of the acts include any noteworthy provisions to prevent fraud.
Playing the data game
New programs in new industries are easy targets for fraudsters, and environmental initiatives — like building renewable energy infrastructure — are still very new in many parts of the country. The Environmental Protection Agency’s enforcement division has brought 16 renewable fuel fraud cases in the last 10 years, with many more cases referred to the Justice Department for criminal prosecution. In 2019, members of a religious sect known as “The Order” pleaded guilty to conspiring with a Los Angeles businessman steal $1 billion in a scheme involving renewable fuel standard credits and related IRS tax credits that involved multiple shell companies. These schemes are about to multiply.
Consider this scenario: An enterprising fraudster decides to create five new shell companies purportedly aimed at building renewable infrastructure. He enlists four friends to put the other companies in their names, so it looks like a competitive group of companies applying for grant funds. They create bogus past performance documents attesting to their qualifications and set to work applying for grant funds in 23 states through the Department of Energy’s Energy Efficiency and Conservation Block Grant, which is now receiving $550 million in funding through the state and community energy program. Despite receiving an influx of billions in new funding, oversight for these programs remains significantly under-resourced.
Overworked and under time pressure, the adjudicators check to be sure all the required documentation is included. The ring uses the same past performance information on all the grant applications, which would be a red flag, but only if the agency had tools to compare text in grant applications across states. Maybe the adjudicators run the companies through a cursory check to ensure they haven’t had any fines or criminal charges levied against them, and since the companies are all new, there are no concerning records that would flag them as suspicious. Because the owners are all different, and the adjudicators don’t have any tools to identify the relationships between the business owners, they have no reason to suspect a ring is involved. With the documents all in order, the adjudicator awards the grants. Once the initial grant payments are deposited into the fraudster’s bank accounts, he closes the accounts and disappears.
AI/ML are key technologies
Preventing fraud is a data game. Bad actors practice “eligibility laundering” across grant programs using rings with small businesses, non-profits and foreign entities all trying to conceal, through deception, their unqualified status within the programs they’ve applied to. Combating these fraudsters requires data, and the tools to use that data quickly and effectively.
The acceleration of machine learning and artificial intelligence tools offers government agencies the ability to identify frauds like the scenario above quickly. Natural language processing text analytics engines can identify duplicate passages in grant applications in seconds. Massive amounts of third-party data can be mined and leveraged to identify past criminal activity and other suspicious indicators related to applicants. Third-party data analysis can also identify patterns indicative of stolen or synthetic identities used in grant and loan applications. Social network analytics can identify the relationships between grant and loan applicants that could indicate the existence of a fraud ring. Device metadata, such as geolocation, can also be mined to identify potential fraud indicators. And all this data analysis can be done in seconds with the tools available today.
If we learned one thing from the pandemic it must be this: Fraud actors have the upper hand. Pandemic relief programs were fleeced in large part due to the lack of data-driven tools and capabilities across federal and state agencies. To protect taxpayer resources, government agencies must invest in the tools needed to fight fraud. For those bills already funded and being implemented, agencies should evaluate their internal capacity to conduct some of the analysis described above, and consider options for redistributing funding to prioritize investment in fraud prevention tools. Going forward, Congress should establish funding for program integrity to accompany large spending bills and direct agencies to establish metrics for preventing fraud actors from stealing the funds. It is simply irresponsible to continue dedicating hundreds of billions of dollars to new spending programs without dedicating the investment in oversight tools to protect the integrity of those funds.
Linda Miller is the founder and CEO of Audient Group, LLC and a nationally recognized expert in fraud risk management. She was a partner at Grant Thornton, LLC where she led the firm’s fraud and financial crimes practice and formerly served as the deputy executive director of the Pandemic Response Accountability Committee (PRAC). She also led the development of the Government Accountability Office (GAO) Framework for Managing Fraud Risks in Federal Programs.
Read more: Commentary
Erik Halvorson is the owner of Archimedes Evaluation Services, LLC, where he supports non-profits and businesses in analytics, program evaluations, and anti-fraud program management. He has 18 years of federal law enforcement experience, including serving as a special agent.
Copyright © 2024 Federal News Network. All rights reserved. This website is not intended for users located within the European Economic Area.
Related Stories
Protecting data from “COVID” fraud
AI/ML are critical to near-peer competition but require support from automated infrastructure
White House $1.6B COVID fraud plan gives federal watchdogs increased staffing