Federal authorities are reportedly using an artificial intelligence tool developed under the Department of Government Efficiency (DOGE) initiative to pinpoint regulations for elimination. By Trump’s inauguration anniversary, government agencies hope to scrap about 50% of the existing federal rules. According to sources, the Department of Housing and Urban Development and the Consumer Financial Protection Bureau (CFPB) are using the “AI Deregulation Tool.” The AI-powered Deregulation Decision Tool will examine roughly 200,000 rules and eliminate what it deems unnecessary, targeting a 50% reduction by next January. The regulations identified by the system will then undergo staff review before removal. The Department of Housing and Urban Development (HUD) has already deployed the tool to assess 1,083 regulatory sections. At the same time, the Consumer Financial Protection Bureau relied on it for all its deregulation decisions. Fields claimed they are still reviewing all options to facilitate deregulation During his campaign, President Trump argued that some federal regulations resulted in inflated prices for several consumer goods. He vowed to launch the “most aggressive regulatory reduction” in US history, including efforts to roll back environmental protections and dismantle the Environmental Protection Agency (EPA). When asked about using AI to deregulate , White House spokesman Harrison Fields said they are reviewing all potential options to fulfill the president’s deregulatory agenda . However, he stressed that no specific plan had been set in stone and that discussions were still in the early stages and involved close coordination with the White House. He added, “The DOGE experts creating these plans are the best and brightest in the business and are embarking on a never-before-attempted transformation of government systems and operations to enhance efficiency and effectiveness.” DOGE’s AI contract reviewer faced issues in June In June, the DOGE AI system mistakenly flagged the Department of Veterans Affairs’ internet contract for cancellation. According to the agency, they gave the service vague and contradictory instructions, hence the mishap. Sahil Lavingia, author of the AI code, had instructed the system to nix anything not directly related to caring for patients. But neither he nor the AI had the know-how to judge that accurately. Lavingia, speaking to ProPublica, acknowledged that he made errors because he didn’t have enough time and tools. He also noted that he understood that the list from the “MUNCHABLE” list was intended to be reviewed by others prior to any decisions. ProPublica’s analysis found that Lavingia’s code also relied on older, generic AI models. The model produced erroneous numbers, often estimating about 1,100 contracts at $34 million each when the real figures were frequently in the thousands, and it struggled to handle entire documents. The analysts also said that using off-the-shelf AI with little understanding of VA operations should never have been on the table in the first place. Waldo Jaquith, a former Treasury Department official appointed under Obama who managed IT contracting, criticized DOGE for prompting the AI system to pull information from easily accessible sources rather than accurate ones, such as publicly available databases like USAspending. He described the approach as lazy; though faster, it’s still less accurate. Cary Coglianese, a University of Pennsylvania professor, also noted that the AI model would need an advanced understanding of medical care, institutional operations, and staffing resources—knowledge it currently lacks. Cryptopolitan Academy: Coming Soon - A New Way to Earn Passive Income with DeFi in 2025. Learn More
An Interview with Fasset
