In his upcoming book, Winton Hall discusses how Code Red: Left, Right, China, and the Race to Control AI, set to be released tomorrow, suggests that AI could significantly improve the detection of large-scale government fraud, like the notorious Minnesota fraud case, potentially saving taxpayers substantial amounts of money.
“AI is particularly adept at identifying waste, uncovering fraud, and highlighting inconsistencies within the federal government,” Hall wrote, noting its ability to process enormous datasets and recognize intricate patterns at unprecedented speeds. “Basically, for the first time in ages, conservatives have a chance to evolve beyond just advocating for limited government. We can actually modernize and boost government efficiency with clever cost-cutting strategies. But, um, where do we even begin?”
According to Hall, AI could serve a crucial role in identifying fraud, much like Elon Musk’s initiative where he gathered a talented group of young tech experts and seasoned professionals to sift through government data, which is seen as a bold move toward significant cost reduction.
One of the visionaries involved, former Tesla engineer Thomas Shedd, headed the General Services Administration’s (GSA) Technology Transformation Services (TTS), which included numerous engineers. He emphasized an “AI-first” approach, crafting an “AI coding agent” for all federal agencies. This agent would analyze government contracts, automate tasks, and help identify unnecessary spending.
Hall elaborates:
The department is working on a comprehensive, secure database to detect duplications and unnecessary spending while improving fraud detection. “What I’m focused on is developing a central system to manage and analyze contracts more effectively. Although it’s not a brand new initiative, it’s something we’ve been preparing for, but now we can potentially build the entire system in-house and quickly. This approach is designed to comply with OPM guidelines regarding data privacy and responsible AI use,” Shedd stated.
According to Dmitry Shevelenko, CEO of AI platform Perplexity, AI could “streamline government operations,” executing “80% of the initial work for target lists faster,” enabling better human decision-making.
Meanwhile, when federal prosecutors started investigating, they targeted abuses involving federal aid. Fraudsters manipulated Minnesota’s Medicaid services, and many individuals charged in these schemes have ties to the state’s Somali community.
In code red, Hall pointed out that the Government Accountability Office (GAO) estimates that fraud costs the federal government between $2.5 billion and $500 million annually. Additionally, since 2023, agencies have reported improper payments totaling approximately $2.8 trillion. Interestingly, the Small Business Administration issued $312 million in loans to children under 11 during the pandemic.
“To be fair, there have been past government attempts using machine learning and data mining that yielded results. Federal law enforcement has made strides in tackling fraudulent schemes that have drained billions from taxpayers,” Hall noted.
He added:
In December 2024, prompted by the House Budget Committee, the Congressional Budget Office (CBO) released its inaugural report on AI. The CBO suggested that AI might enhance governmental efficiency in tax collection and distribution through transfer payments, and that effectively using AI could cut down on inappropriate payments in significant mandatory spending programs like Medicare, Medicaid, and Social Security. Addressing this is crucial; improper Medicaid payments are estimated between $543 billion and $1.1 trillion over the past decade.
Hall continued, referring to code red, that the Office of Personnel Management (OPM) indicated that generative AI could enhance federal workers’ output, increasing creativity, efficiency, and productivity.
In code red, Hall addresses myriad topics surrounding AI, including its implications for elections, the economy, and even family life. Importantly, he views AI not as an inherently good or bad force, but rather as a tool needing careful management to align with American values.
Sen. Marsha Blackburn (R-Tenn.) recognizes Hall as one of the 100 most influential figures in the AI field and calls code red a “must read.” Hall is “uniquely qualified” to discuss how to maximize AI’s vast potential while safeguarding against exploitation of vulnerable populations like children, creators, and conservatives. Michael Shellenberger, an acclaimed investigative journalist, describes code red as “insightful” and “disturbing” and notes it sparks essential conversations regarding resisting Big Tech’s authoritarian agenda before it’s too late.
