The tool has already been used to complete “decisions on 1,083 regulatory sections” at the Department of Housing and Urban Development in under two weeks, according to the PowerPoint, and to write “100% of deregulations” at the Consumer Financial Protection Bureau (CFPB). Three HUD employees — as well as documents obtained by The Post — confirmed that an AI tool was recently used to review hundreds, if not more than 1,000, lines of regulations at that agency and suggest edits or deletions.

The tool was developed by engineers brought into government as part of Elon Musk’s DOGE project, according to two federal officials directly familiar with DOGE’s work, who, like others interviewed for this story, spoke on the condition of anonymity to describe internal deliberations they were not authorized to discuss publicly.

End of carousel

Conservatives have long argued that the federal government issues far too many regulations that constrain economic growth and hurt the private sector. Many liberals have emphasized that there are reasons federal regulations are in place, such as protecting the environment and ensuring food safety.

Asked about the AI-fueled deregulation, White House spokesman Harrison Fields wrote in an email that “all options are being explored” to achieve the president’s goal of deregulating government. Fields noted that “no single plan has been approved or green-lit,” cautioning that the work is “in its early stages and is being conducted in a creative way in consultation with the White House.”

Fields 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.”

Follow Trump’s second term

No decisions have been completed on using AI to slash regulations, a HUD spokesperson said. The agency is having “ongoing discussions” to consider how to make government more efficient. “We are not disclosing specifics about how many regulations are being examined or where we are at in the broader process,” the spokesperson said, adding, “the process is far from final.”

The spokesperson continued: “The intent of the developments is not to replace the judgment, discretion and expertise of staff but be additive to the process.”

One former member of DOGE, which stands for Department of Government Efficiency, wrote in a text message that the team did everything it could to come up with legal and technological solutions to repeal as many regulations as possible within Trump’s term.

“Creative deployment of artificial intelligence to advance the president’s regulatory agenda is one logical strategy to make significant progress in that finite amount of time,” wrote James Burnham, who served as chief attorney for DOGE and is now managing partner at King Street Legal.

The proposed use of AI to accomplish swift, massive deregulation expands upon the Trump administration’s work to embed AI across the government — using it for everything from fighting wars to reviewing taxes. And it dovetails with the administration’s aim to unwind regulations government-wide, even without AI. But it’s unclear whether a new, untested technology could make mistakes in its attempts to analyze federal regulations typically put in place for a reason.

On Jan. 31, Trump issued an executive order to “unleash prosperity through deregulation,” which required agencies to repeal 10 rules for every new rule issued. Since then, some departments have engaged in what almost appears to be a competition to cut. In May, the Transportation Department declared it had deleted 52 regulations and more than 73,000 words from the Federal Register. This month, the Labor Department announced plans to nix more than 60 regulations.

Still, Republicans have grown frustrated by the relatively slow pace of deregulatory actions. During the first six months of Trump’s first term, his administration cut costs by about $550 million and paperwork hours by 566,000, according to the American Action Forum, a center-right think tank that tracks regulations. Through July of this year, the Trump administration has achieved nearly all its cost reductions by repealing one rule regarding what businesses must report about their ownership ties. Without that, the Trump administration would have increased regulatory costs by $1.1 billion and paperwork hours by 3.3 million, according to the think tank.

“They’re way behind where they were in 2017 on the numbers, no question about it,” said Doug Holtz-Eakin, president of the American Action Forum and former director of the nonpartisan Congressional Budget Office. “I thought this was going to be something they crushed because they did so in 2017. I’ve been baffled by this.”

The AI tool is intended to massively accelerate the deregulation process, with every federal agency able to develop a list of regulations to eliminate in less than four weeks, according to the PowerPoint. The agencies are supposed to finish their lists by Sept. 1, and this month, DOGE is supposed to start training staff at agencies on how to use the AI tool, the PowerPoint states.

Read the full PowerPoint here.

While DOGE had pushed earlier this year to take a larger role in the deregulatory effort, the Musk-led team was frequently rebuffed by agency employees who worried about outsourcing decisions and their authorities, according to three people who have participated in deregulatory conversations at the White House and the agency level who spoke on the condition of anonymity to share private conversations. Federal officials also questioned whether DOGE had the subject matter expertise to comb through highly technical regulations and find appropriate targets for cuts, the people said.

As DOGE’s influence waned following Musk’s departure, the administration has remained focused on Trump’s deregulatory order, the people said. White House staff are also using internal trackers to monitor how quickly agencies are paring regulations, while leaders at every major agency are meeting regularly to discuss how quickly they can meet Trump’s ambitions and which cuts “count” toward the president’s order, according to the people.

In some cases, DOGE’s campaign to fire federal workers and dramatically shrink the federal workforce has hampered the deregulatory effort, the three people said.

“The White House wants us higher on the leader board,” said one of the three people. “But you have to have staff and time to write the deregulatory notices, and we don’t. That’s a big reason for the holdup.”

Trump officials have tried to use AI to roll back regulations before. At the Department of Health and Human Services, a 2020 “Regulatory Clean Up Initiative” drew on an AI tool to identify and remove archaic language, defunct federal provisions and outdated terms from federal rules.

Trump has pushed the limits of the Administrative Procedure Act, which governs repealing federal regulations, most notably through an executive order ending a rule that restricted the water flow of showerheads. It is unclear if courts will allow the administration to void rules. Meanwhile, private-sector companies tend to be uncomfortable ignoring a rule that was illegally repealed, said Nicholas Bagley, an administrative law expert at the University of Michigan.

“There’s been some flashy sideshow efforts to avoid the legal strictures, but in general, they don’t stick,” Bagley said of Trump’s unilateral efforts to cut regulations.

DOGE officials may be concerned about the legality of the AI tool. One page of the slideshow says four people identified as “DOGE lawyers” — Burnham, Austin Raynor, Jacob Altik and Ashley Boizelle — each “vetted and endorsed” the AI deregulation tool. Raynor, Altik and Boizelle could not be reached for comment.

Federal regulations, as they stand now, can be divided into three categories, the PowerPoint says: 50 percent are not required by law, 38 percent are statutorily mandated and 12 percent are “Not Required but Agency Needs.” By ending the rules that are both unnecessary by law and to agency operations, the PowerPoint states, the government could recover $3.3 trillion a year. But the PowerPoint also suggests it would take 3.6 million “man-hours” to nix 100,000 regulations under the current system. It is not clear how the PowerPoint’s authors arrived at these figures.

That’s where the AI tool comes in, the PowerPoint proposes. The tool will save 93 percent of the human labor involved by reviewing up to 500,000 comments submitted by the public in response to proposed rule changes. By the end of the deregulation exercise, humans will have spent just a few hours to cancel each of the 100,000 regulations, the PowerPoint claims.

The PowerPoint lists two case studies as examples of how well its AI tool can work, detailing recent efforts to slash regulations at HUD and CFPB.

CFPB did not respond to questions. The Post was not able to independently confirm the use of AI at the agency.

At HUD, efforts to use AI to kill regulations began three months ago, according to three employees familiar with the matter and emails obtained by The Post. A message sent to some of the agency’s Public and Indian Housing staff on April 18 announced a “DOGE team” would be “learning how AI will be able to analyze all PIH regulations looking for and flagging discrepancies between them and the underlying statute.”

“This is a major effort,” the email continued. “We are working with the lawyers to simplify the [Administrative Procedure Act] process … use AI for drafting, and use AI for complying notices in the future.”

The overall goal, the email noted, was to deploy AI to reduce the time staff had to spend on deregulation. Another document, signed “HUD DOGE Team” and sent to staff, detailed how DOGE team members wanted federal staffers to engage the AI tool. Staffers were supposed to look over the tool’s recommendations for proposed regulatory eliminations and mark whether they agreed, disagreed or believed deletions should go further.

One HUD employee who participated in this process said the AI tool made several errors. It delivered an analysis saying those who drafted various agency regulations had misunderstood the law in several places, said the employee, who spoke on the condition of anonymity to reveal internal conversations. But the AI tool was sometimes wrong, the employee said.

“There were a couple places where the AI said the language was outside of the statute,” the employee said, “and actually, no — the AI read the language wrong, and it is actually correct.”

After its tryout at HUD, the AI deregulation tool is supposed to deploy across the rest of government in coming months, according to the DOGE PowerPoint.

Over the next five months, agencies will work with the AI tool to identify regulations to kill, respond to public comments about the proposed deletions and submit formal deregulation proposals, the PowerPoint says. The goal is to wrap everything up and “Relaunch America on Jan. 20, 2026,” the PowerPoint states.

Lauren Kaori Gurley and Ian Duncan contributed to this report.