Last Wednesday, the Chief Information Officer Office at the Management and Budget Office hosted a video call to discuss an artificial intelligence tool used to reduce federal regulations, which the office called Sweetrex deregulation. The tool, which is still being developed, is created to identify the sections of the regulations that are not requested by the statute, therefore accelerating the process for the adoption of updated regulations.
The development and launch of what is formally called Sweetrex Deregulation at Plan Builder, or Sweetrex Daip, aims to help achieve the objectives established in President Donald Trump “To trigger prosperity through deregulation” executive orderwhich aims to “promote prudent financial management and alleviate unnecessary regulatory charges”. Industrial scale deregulation is a fundamental objective established in project 2025, the document that served by playbooks for the second Trump administration. The so -called department of government efficiency (doge) has also estimated that “50 percent of all federal regulations can be eliminated”, according to a July 1, 2025, Powerpoint presentation obtained from The Washington Post.
To this end, Sweetrex was developed by Doge members operating by the Department for Housing and Urban Development (HUD). The plan is to launch it to other US agencies. The members of the call included the staff of the whole government, including the Environmental Protection Agency, the State Department and the Federal Depaction Insurance Corporation, among others.
Christopher Sweet, an affiliate of Doge who was initially presented to colleagues as a “special assistant” and who until recently was a third year student at the University of Chicago, transmitted the call and was identified as the main developer of Sweetrex (therefore, his name). He told colleagues that Anthropic and Opens tools will be increasingly used by federal workers and that “many productivity increases will come from the tools that are built around these platforms”. Sweet said that for Sweetrex, they are “mainly using the Google models family, therefore mainly twins”.
Neither Sweet nor Ora immediately replied to Wired’s commentary request. The Hud press office responded only to say that the request was “in the revision phase”. Google has not yet responded to a commentary request.
Previously, Wired reported on the output of an AI for deregulation tool at HUD. A calculation sheet detailed how many words could be eliminated by the individual regulations and has given a percentage figure that indicates how much the regulations were not conforming; How was that percentage calculated was not clear. At the time, Sweet did not respond to a commentary request and a Hud spokesman said that the agency does not comment on the individual staff.
Wednesday’s call with Sweet was Scott Langmack, a senior consultant affiliated to Doge at the Hud and, second His LinkedIn profileThe Coo of Technology Company Kukun. (Wired previously, he reported having access to application level to critical HUD systems; Kukun is a company of Proptech who is, according to his website“On a long -term mission to aggregate the most difficult to find data.”) While Sweet guided the development side of Sweetrex, Langmack said he was taking stock of the demonstration of the instrument for different agencies and putting them on his benefits. For example, he said that the tool is able to reduce the time spent to review and propose regulations modified for months to a few hours or days.
Be First to Comment