The Digital Coup: Inside Elon Musk’s DOGE Takeover of the US Government

When Washington’s most decorated senior civil servants sipped cocktails in the Capital Hilton’s presidential ballroom, fretting over table placements and their rank in the unspoken pecking order between a sitting US senator and the UAE ambassador, Elon Musk sat hunched over his phone, laughing to himself.

Very few guests at the January 25 Alfalfa Club banquet in Washington, DC, knew what Musk already knew: a team of senior executives and young, die-hard Musk loyalists was preparing to seize control of top floors of a nearby federal building. Under armed guard, they would sleep on mattresses fitted with biometric sensors tracking body temperature and breathing rate, all while racing to overhaul the federal government’s entire underlying code base — or better yet, scrap the whole system entirely.

Musk never cared for formalities, but he’d dressed up for the occasion. Founded in 1913, the Alfalfa Club exists for one purpose only: to host an annual banquet where Washington’s most powerful political figures can hobnob with the nation’s top business leaders. Membership is capped at roughly 200, and new "sprouts" (as the club calls incoming members) are only admitted when a sitting member dies. That evening, new inductees included Supreme Court Justice Elena Kagan and Nasdaq CEO Adena Friedman, joining existing members like Microsoft CEO Satya Nadella and Senator Kirsten Gillibrand. Musk attended only as a guest.

The club’s social organizers stuck to tradition, letting elections and constitutional norms dictate the seating chart. The head table was reserved exclusively for Alfalfa members who held public office. Musk, the widely rumored leader of the new Department of Government Efficiency, was seated across the room on the opposite end. He spent most of the dinner glued to his phone — if on-the-night rumors were to be believed, he was on a call with President-elect Donald Trump. By that point, Musk was closer to Trump than he had ever been. He told friends he was already crashing in federal government buildings, and would soon move next door to the White House, taking up residence in the Secretary of War Suite inside the Eisenhower Executive Office Building. He’d even already had his custom gaming rig installed there.

While Musk sat in the Hilton ballroom, his operatives — led by one of his most trusted lieutenants — had already secured access to internal systems at the Office of Personnel Management, the federal agency that manages human resources for roughly 2.2 million career civil servants across the government. Many of these operatives would soon fan out to occupy offices across virtually every federal agency. Their ranks include:

• Akash Bobba, a UC Berkeley graduate and former intern at Palantir, the defense contractor co-founded by Peter Thiel

• Edward Coristine, a 19-year-old engineer who goes by the online alias "Big Balls"

• Nikhil Rajpal, a 30-something engineer who worked at Twitter during Musk’s acquisition, where he once pitched a plan to auction off inactive usernames to the highest bidder. As an undergrad at UC Berkeley, Rajpal led a libertarian student group that adopted the Latin motto Futuate cohortem urbanam — roughly translated to "Fuck these city dwellers."

In Musk’s view, Washington was not just bloated — it was buggy, in need of a hard fork and a full sunset of all legacy systems. His strike teams of young engineers would burrow into the government’s byzantine, decades-old bureaucratic systems and delete anything they deemed unnecessary. Their core mission was to help Trump slash the federal budget to the bone. At the Alfalfa dinner, Musk turned to the people at his table: Can you believe we’re spending taxpayer money on condoms? The others shook their heads. Musk looked back at his phone, then added: What if we cut all federal grants to NGOs entirely?

In the days and weeks that followed, DOGE swept through one part of the federal government after another. The Departments of Agriculture, Defense, Education, Health and Human Services, Homeland Security, Housing and Urban Development, Justice, and Veterans Affairs; the Federal Aviation Administration, General Services Administration, Social Security Administration, and National Oceanic and Atmospheric Administration; the Consumer Financial Protection Bureau and Internal Revenue Service; the US Agency for International Development and the Centers for Disease Control and Prevention; the National Park Service and the National Science Foundation — all fell under DOGE’s control. An estimated tens of thousands of federal employees were effectively fired or chose to resign. "This is a digital coup," one USAID source told WIRED at the time.

| Got a Tip? |

|---|

| Are you a current or former government employee who wants to talk about what's happening? We'd like to hear from you. Using a nonwork phone or computer, contact the reporters securely on Signal at makenakelly.32, velliott88.18, dell.3030, DavidGilbert.01, Kateknibbs.09, or dmehro.89. |

Alongside its push for budget cuts and staff reductions, DOGE also gained access to untold terabytes of sensitive personal government data. Trump gave Musk and his operatives carte blanche access to any unclassified system they wanted to tap. One of their first targets was a database previously breached more than a decade earlier by alleged Chinese cyberspies, which holds investigative files on tens of millions of US government employees. Other data stores opened to DOGE include federal workers’ tax records, biometric data, private medical histories (such as treatment for drug and alcohol abuse), cryptographic keys for restricted areas at federal facilities nationwide, personal testimonies from low-income housing recipients, and granular location data for particularly vulnerable children.

Multiple government finance, IT, and security specialists told WIRED none of this sensitive data appears relevant to Musk’s stated goal of identifying waste and fraud. But by treating the entire US government as one giant dataset, experts say DOGE could help the Trump administration accomplish an unstated second goal: aggregating every piece of information the government holds on any individual — from a career civil servant to an undocumented immigrant — into a single, easily searchable central repository.

WIRED spoke with more than 150 current and former federal employees, experts, and Musk supporters across more than 20 agencies to expose DOGE’s inner workings. Most sources requested anonymity to speak candidly about DOGE’s actions and future plans.

Musk and Trump’s partnership was cemented on July 13, 2024, when a would-be assassin came within inches of killing the former president at a rally in Butler, Pennsylvania. Musk was galvanized by the iconic photo of Trump, blood streaming down his face, raising a fist and shouting "Fight, fight, fight" for cameras. The image quickly became a viral meme — Musk’s preferred form of communication. He endorsed Trump that same day, and reoriented his recently launched super PAC to get the former president reelected.

The following month, during a live discussion on X, Musk floated the idea of leading a "government efficiency commission" for a second Trump term. Trump’s response was enthusiastic. "You're the greatest cutter," he said admiringly.

Two years earlier, after Musk purchased Twitter in a chaotic blitz of last-minute paperwork and nine-figure money transfers, he cut roughly 80% of the company’s staff, closed at least a dozen international offices, and rolled back Twitter’s content moderation policies in the name of free speech. He demanded change at such breakneck speed that one of his lieutenants, Steve Davis, ended up sleeping at Twitter’s San Francisco headquarters with his partner and their newborn baby.

In Washington, Musk estimated his team could cut "nearly $2 trillion" from the annual federal budget. After setting aside non-discretionary spending such as Medicare, Social Security benefits, and national debt interest payments, that $2 trillion figure is almost equal to the entire remaining federal budget. Put another way: Musk was effectively proposing cutting almost everything else, from foreign aid to housing subsidies, from national park maintenance to basic weather data collection, from investigations into predatory lenders to the operation of US air traffic control.

After Trump won the election, he announced that Musk and former Republican presidential candidate Vivek Ramaswamy would co-lead DOGE. The announcement kicked off a stealth recruiting process led by Davis, the same executive who slept at Twitter HQ. Musk envisioned a team of ultra-high-IQ individuals joining him in Washington for an 18-month, 80-hour-a-week hackathon on the US government.

The DOGE leadership set up camp on the eighth floor of SpaceX’s Washington, DC office, commandeering multiple conference rooms to hold meetings and interviews with DOGE applicants, according to a person familiar with the process. One question was asked of every candidate: Who did you vote for in 2024?

Among Davis’ early recruits was Zsombor (Anthony) Jancso, a mid-20s San Francisco-based engineer and former Palantir employee. After Palantir, Jancso worked on a project called Accelerate X, which billed itself as "a modern OS for government" with solutions "delivered in days." His co-founder, MIT-educated engineer Jordan Wick, joined DOGE as well.

A few weeks after the 2024 election, an online handle linked to Jancso reached out to a group of participants in a US Space Force AI challenge. The poster said they were looking for "hardcore engineers" and instructed applicants to send their GitHub or LinkedIn profiles to @DOGE on X, then reply privately with their own X handle. (To do this, applicants needed a paid X Premium subscription.) Not long after, the same handle posted in a group for Palantir alums: "This is a historic opportunity to build an efficient government, and to cut the federal budget by 1/3."

Twenty-three-year-old engineer Luke Farritor quickly joined DOGE’s recruitment effort. The son of a University of Nebraska-Lincoln academic and a physician, Farritor was homeschooled, and his childhood bedroom was filled with books by James Baldwin and Jordan Peterson. In college, he landed a prestigious internship at SpaceX and wore his SpaceX t-shirt almost constantly. Farritor gained nerdy prominence for his work using machine learning to decipher an ancient papyrus charred by the same volcanic eruption that buried Pompeii. He has said his work received support from a $2 million grant from the Musk Foundation, and Musk publicly backed the project on X. The papyrus project also earned Farritor a Thiel Fellowship last spring — a $100,000 grant funded by Peter Thiel to encourage bright young people to drop out of college, which Farritor promptly did.

On December 5, Farritor posted in a Discord group for SpaceX interns, noting that DOGE was looking for "skilled software engineers (and ops people) at any career stage who are willing to work for ~6mo in person in DC. Paid." He added: "We’re going to fix the government!"

Meanwhile, Musk was spending time at Mar-a-Lago getting a crash course in American civics, taught by a rotating cast of Washington bureaucrats, venture capitalists, and right-wing X influencers. One of Musk’s advisors was Antonio Gracias, a private equity investor and early Tesla backer, who later summed up what they learned on a podcast: "A department just basically asks for money from Treasury and they send it out."

Of course, the truth is far more complex. Before Treasury cuts a check, payment authorization must pass through a web of custom-built technical systems built over decades to ensure money appropriated by Congress is spent properly by the executive branch. The system is inefficient by design, stacked with layer after layer of backstops to prevent misuse. To Musk, that only meant it was ripe for disruption. (Gracias would later go on to become a DOGE "IT specialist" at the Social Security Administration.)

Even Musk eventually acknowledged his original $2 trillion target was overly aggressive. Economist Dean Baker told WIRED that level of cutting would deliver a shock to the US economy on par with the 2008 financial crisis, with the potential to push unemployment into double digits. By mid-January, Musk walked the target back to $1 trillion — a sum still larger than the entire nondefense discretionary budget. He began targeting individual agencies, and was especially eager to eviscerate USAID, the federal government’s primary foreign aid body. Right-wing conspiracy theories Musk later reposted on X frame USAID as a criminal enterprise, a tool of the "deep state’s" woke agenda.

As Trump’s second inauguration drew near, Musk and Ramaswamy’s visions for DOGE diverged sharply. Ramaswamy advocated for an incremental approach: push for legislative changes that would eventually cut off spending at the source. Musk refused to wait. Increasingly, his plans aligned with a framework incoming Vice President JD Vance laid out in a 2021 podcast interview: fire all midlevel career civil servants and replace them with Trump loyalists. Just before Trump took office, Musk’s vision won out, and Ramaswamy left the organization.

Trump officially established DOGE on the afternoon of his inauguration. He effectively folded it into the existing US Digital Service, an Obama-era agency created to bring private-sector tech talent into short-term civil service roles. Now the "D" in USDS stood for "DOGE." The order also created a temporary subsidiary body, the US DOGE Service Temporary Organization, set to expire on July 4, 2026. This structure allowed DOGE to bring in special temporary government employees: people who serve for a limited time before returning to the private sector, and critically, are not subject to the same transparency requirements as regular career civil servants.

After inauguration, the DOGE brain trust vacated their SpaceX DC offices. An employee who worked there says they left a men’s bathroom "trashed," with one urinal "filled with gum and Zyn."

Daisy Kid Henderson first heard from DOGE around 9:30 pm on Inauguration Day, when she received an email inviting her to a meeting with anonymous staffers the following afternoon. Henderson, a fiery 28-year-old software engineer at USDS, describes herself as an "eternal optimist." Even so, the lack of last names on the invite left her unsettled.

Henderson, who is based in Denver, had worked at USDS since January 2024. She joined after seven years at Comcast, where she worked on everything from quantum computing to robotics to building an audio algorithm that filters sounds triggering distress for people with PTSD or autism. For Henderson, USDS was a chance to do good while reevaluating her career. "It was my dream job," she recalls. "When I was hired on, it was even better than I imagined." Henderson oversaw three generative AI pilot programs at the Department of Homeland Security. One, built for the Federal Emergency Management Agency, helped local officials plan and apply for disaster aid funding. She also partnered with the Food and Nutrition Service to help multiple states deliver assistance to low-income families faster.

By the time Henderson’s meeting with DOGE rolled around, "they had been doing these meetings all day nonstop," she recalls. DOGE "had four or five interviewers and worked in shifts," likely to churn through all 200 legacy USDS employees as quickly as possible, Henderson says. According to another USDS staffer, when someone asked the interviewers for their last names, the reply was: "This is a one-way conversation. We won't be answering any questions."

One question DOGE asked every candidate: What do you think about DOGE?

Henderson tried to stay positive as her call with two young DOGE operatives began. They introduced themselves only by first names: Cole and Ram. (They were Cole Killian, a 24-year-old who attended McGill University, and Nikhil Rajpal, the former libertarian student leader and ex-Twitter engineer.) As the men asked Henderson about her projects, they were particularly interested in her AI work for FEMA and her technical skills. Then the questions took a strange turn, centering around "who the underperformers are at USDS," Henderson recalls, and what skills she brought to her role. At one point, Killian simply got up and walked out of the room without a word. Rajpal carried on as if nothing had happened. Afterward, Henderson learned her colleagues had been asked whether they deserved to keep their jobs.

Similarly unusual meetings played out at the General Services Administration, another early DOGE target. If OPM is the federal government’s HR department, GSA is its combined operations and IT department rolled into one. It oversees more than 1,000 federally owned buildings, from generic office parks to iconic architectural landmarks, along with hundreds of thousands of government vehicles and tens of billions in annual government procurement.

Just as it would do at agencies across the government, DOGE’s takeover of GSA happened out of public view: no official announcements, just unexpected calendar invites from unknown senders and unfamiliar names popping up in internal directories. The sixth and seventh floors, home to the administrator’s "A-suite" offices, were locked down and restricted to most employees. Employees could no longer just badge in through the turnstile; now they had to pass through metal detectors and have their belongings x-rayed.

During that first week, GSA employees caught a glimpse of a whiteboard in a large vacant room with three items written on it. No rank-and-file employee knew who wrote it, or what it meant.

As Musk’s top team at GSA, DOGE installed Nicole Hollander — Steve Davis’ partner, the other parent of the newborn that slept with Davis at Twitter HQ. Former Tesla software engineer Thomas Shedd became director of GSA’s Technology Transformation Services, which operates dozens of critical systems used across all federal agencies, including Login.gov, Cloud.gov, and the Federal Procurement Data System (the public database that lists all unclassified government contracts above the micropayment threshold). DOGE’s strike team at GSA included Coristine, the young engineer known as "Big Balls," and Farritor, the papyrus-decoding wunderkind. Another young DOGE recruit, Ethan Shaotran, also got a GSA email account and A-suite clearance. Shaotran recently served as president of Harvard’s mountaineering club, and connected with DOGE at a hackathon for xAI, another Musk-owned company.

It soon became clear DOGE wanted GSA to roll out one specific product: an AI chatbot that could plug directly into the agency’s main data portal, the Enterprise Data Solution. The tool would let a handful of DOGE technicians ask questions in plain language and pull answers from vast stores of government data. (It remains unclear how this complies with GSA’s Internal Data Sharing Policy, which requires supervisor approval for requests for certain types of controlled unclassified information.) For DOGE operatives unfamiliar with GSA’s complex systems, this might have seemed like a quick build — especially if they started with an off-the-shelf large language model like Claude, Gemini, or Llama.

But GSA’s veteran engineers knew the project DOGE wanted was far more complex than it looked. The Enterprise Data Solution is a maze of disconnected databases, analytical tools, and machine learning systems,