“Having the best spies, the best collection systems, and the best analysts will not help an intelligence service if it leaks like a sieve,” the former CIA speechwriter Charles E. Lathrop remarked in The Literary Spy, a book of quotations about espionage that he compiled. Lathrop, who wrote under a pseudonym, was making a point about counterintelligence—the flushing out of enemy spies and leakers who might compromise a spy agency’s precious secrets. Counterintelligence, Lathrop observed, “is the kidneys of national security: necessary, but unheralded until something goes wrong.”

These days, something looks to have gone very wrong—with the kidneys and maybe with the brain, too.  

To protect secrets, people who will be handling classified information or assuming positions of trust within intelligence agencies are vetted, often by law-enforcement agents, who interview friends and co-workers, review travel histories, and analyze financial information to determine whether someone might make an attractive recruit for a foreign intelligence service. Perhaps he’s in debt and would be willing to sell sensitive information. Or maybe she harbors some allegiance to a hostile country or cause and might be willing to spy for it. Looking for these red flags is counterintelligence 101, an imperfect, laborious, and invasive process that American presidents of both major parties have nevertheless accepted as the cost of doing intelligence business.

[David Deming: DOGE is failing on its own terms]

But the legion of Elon Musk acolytes who have set up shop inside federal agencies in the past few weeks do not appear to have been subjected to anything approaching rigorous scrutiny. President Donald Trump has also nominated to key national-security positions people whose personal and financial histories contain at least caution flags. This deviation from past practice has created a new kind of counterintelligence predicament, officials and experts have told me. Rather than staying on high alert for hidden threats, the counterintelligence monitors have to worry about the people in charge.

The public knows very little about how, or if, staff at the new Department of Government Efficiency that Musk runs were vetted before they obtained access to the Treasury Department’s central payment system or the files of millions of government employees at the Office of Personnel Management. These two databases could help U.S. adversaries uncover the identities of intelligence officers and potentially their sources, people with knowledge about how the systems are set up told me.

Precisely what the DOGE teams are doing with this information, whom they’ve shared it with, and whether they have adequately protected it from falling into the wrong hands remains unknown. But the risks posed by this direct access to the government’s central nervous system are entirely foreseeable.

“The fact that people are getting access to classified and personally identifiable information who are not being vetted by our national-security system means it is more likely that there are going to be damaging leaks,” Tim Naftali, a counterintelligence expert and presidential historian at Columbia’s School of International and Public Affairs, told me.

Why would President Trump, who is the ultimate arbiter of who gets to see classified information, take such risks? One answer is rooted in his historic distrust of the FBI, whose agents traditionally conduct background investigations of senior administration officials as they assume their posts. Trump views the bureau as a hotbed of disloyal conspirators. During the presidential transition, he reportedly resisted efforts to allow FBI background checks, and how thoroughly members of his administration were vetted, if at all, is still not clear.

Animus and mistrust likely guide the president’s decisions here. He has publicly seethed at the agents who searched his Florida home, as part of an investigation that led to felony charges for mishandling national-security information after he left office. The agents who worked on that case are assigned to a counterintelligence squad at the FBI’s Washington field office, and the White House is trying to fire them. These agents routinely investigate threats to U.S. national security, and removing them would at least temporarily stall their efforts.

“In his dark passion for retribution, Trump is making his own government, which is our government, more vulnerable to adversarial penetration,” Naftali said.

Security risks now pervade the federal government, thanks largely to a cadre of youngsters, some barely out of high school, whom Musk has deployed inside federal agencies, ostensibly to identify wasteful government spending. In addition to the Treasury Department and the Office of Personnel Management, DOGE agents have reportedly accessed information networks at the State Department, the Centers for Disease Control and Prevention, the Centers for Medicare & Medicaid Services, the Commerce Department, the Education Department, and the Energy Department, among others. Musk has further plans to send teams to other major organizations, including the Pentagon.

[Read: The government’s computing experts say they are terrified]

As his teams fan out, the kidneys of counterintelligence are backing up.

At Treasury, a security team warned that DOGE employees’ access to a central government payment network presents an “unprecedented insider threat risk,” The Washington Post reported last week. The government defines an insider threat as “someone with regular or continuous access” to a computer system who could exploit the information for criminal purposes, leak it to unauthorized parties, or sell it to a foreign government. Edward Snowden, the government contractor who disclosed classified information about NSA surveillance to journalists and who now lives in Russia, is the classic modern example.  

Two intelligence officials told me that the Treasury system, which processes more than $5 trillion in payments each year, contains sensitive national-security information. It could be used to uncover the identities of U.S. intelligence officers—who are after all paid from the Treasury—as well as people or organizations who are paid to spy on behalf of the United States.

These names are not explicitly identified as intelligence assets in the Treasury network, but an adversary with the time and know-how could use the Treasury data, possibly in concert with other information, to discover classified identities, the officials indicated. According to the Post, a senior career official at the department raised such concerns in a letter to Treasury Secretary Scott Bessent. The official recommended some unknown mitigating steps that Bessent reportedly approved.

At the Office of Personnel Management, DOGE employees gained access to information, including addresses and salary history, about Treasury and State Department employees working in “sensitive security positions,” the Post also reported. Personnel data are another puzzle piece that could allow an adversary to identify who works for the intelligence community, and potentially in what country they’re stationed.

“Little pieces of information matter a lot when they’re put together with other little pieces of information,” Joel Brenner, who was in charge of U.S. counterintelligence policy under Presidents George W. Bush and Barack Obama, told me. This is standard intelligence tradecraft. “That’s how we do it. That’s how every intelligence service does it,” Brenner said.

The Office of Personnel Management is not known for its counterintelligence prowess. A decade ago, Chinese hackers breached the agency’s computer networks and stole the records of millions of U.S. government employees, in one of the great espionage coups of recent history. As I reported at the time, officials had earlier resisted a plan to merge a system known as Scattered Castles, which contained the records of intelligence-agency personnel and others who held security clearances, with OPM’s system, fearing exposure in just this scenario.

Their concerns proved prescient, and today, Scattered Castles remains segregated from OPM’s systems—fortunately, given recent reports that Musk’s team has connected its own server to OPM’s systems, which could open a gateway for foreign hackers to again burrow in.

Yet intelligence-personnel records may still be at risk. Last week the CIA sent OPM a list of names of new CIA officers via an unclassified email, people familiar with the matter told me. The CIA sent only the officers’ first names and the first initial of their last names. But even those fragments of information could be useful to foreign spies.

Over the weekend, a former senior CIA official showed me the steps by which a foreign adversary who knew only his first name and last initial could have managed to identify him from the single line of the congressional record where his full name was published more than 20 years ago, when he became a member of the Foreign Service. The former official was undercover at the time as a State Department employee. If a foreign government had known even part of his name from a list of confirmed CIA officers, his cover would have been blown. The cover of a generation of young intelligence officers now appears to depend on whether Musk’s DOGE kids are, with no obvious experience in such matters, properly handling and protecting the information that the CIA sent them.  

How trustworthy are Musk’s employees? Early reports suggest that if they had been subject to traditional background checks, which they apparently were not, some of them would have had trouble passing. One standout in this regard, Edward Coristine, a 19-year-old DOGE member who has used the online handle “Big Balls,” was fired from an internship after he was accused of sharing proprietary information with a competitor, Bloomberg reported. After he was dismissed, the former intern bragged on an online chat platform that he “had access to every machine” and could have deleted crucial data from the company’s servers. “I never exploited it because it’s just not me,” Coristine reportedly wrote. This is the textbook definition—indeed, the U.S. government’s definition—of an insider threat.

The cybersecurity journalist Brian Krebs has written that Coristine was affiliated with a community of chat channels “that function as a kind of distributed cybercriminal social network.” Coristine, who was first identified not in a government announcement but by investigative reporters at Wired, founded a company that “controls dozens of web domains, including at least two Russian-registered domains,” the publication reported. Coristine has recently been named a senior adviser at the State Department, according to the Post.

[Read: If DOGE goes nuclear]

Government computer-security experts are worried that DOGE members could corrupt vital technology systems. “Musk and his crew could act deliberately to extract sensitive data, alter fundamental aspects of how these systems operate, or provide further access to unvetted actors,” my colleagues wrote in The Atlantic last week. An insider need not even behave maliciously to cause havoc. DOGE agents, who are overwhelmingly young with little professional experience or familiarity with older government systems, “may act with carelessness or incompetence, breaking the systems altogether. Given the scope of what these systems do, key government services might stop working properly, citizens could be harmed, and the damage might be difficult or impossible to undo.”

The counterintelligence risks don’t extend only to unchecked young people with the keys to the government’s kingdoms of data. Some of Trump’s Cabinet nominees—including those for two national-security positions—raise classic red flags.

According to his financial disclosure forms, Kash Patel, Trump’s nominee to run the FBI, was paid $25,000 last year by a film company owned by a dual U.S.-Russian citizen that has made programs promoting “deep state” conspiracy theories pushed by the Kremlin, the Post reported. Receiving money from a foreign government is a basic risk factor because it raises questions about whether a government employee’s favor or influence can be bought.

The resulting six-part documentary appeared on Tucker Carlson’s online network, itself a reliable conduit for Kremlin propaganda. In the film, Patel made his now infamous pledge to shut down the FBI’s headquarters in Washington and “open it up as a museum to the ‘deep state.’” The FBI is one of the Russian intelligence services’ main targets for espionage.

On his disclosure forms, which were made public only after he testified in his Senate confirmation hearing, Patel describes the payment as an “honorarium.” That term traditionally implies a nominal or even negligible sum of money, which this was not. He also listed consulting work for clients that include the Qatari embassy and said that he would keep his stock in the Cayman Islands–based parent company of the clothing brand Shein, which was founded in China.

According to his financial disclosure forms, Robert F. Kennedy Jr., Trump’s nominee to run the Health and Human Services Department, is saddled with up to $1.2 million in credit-card debt. Owing money is another risk factor because it might induce people to accept funds in exchange for sensitive information. Investigators examine bank records, credit-card statements, and other financial documents to determine how much debt a security-clearance applicant carries and its proportion to his level of income.

Allegiance or even sympathy to a hostile power is yet another warning sign. Tulsi Gabbard, Trump’s director of national intelligence, has drawn widespread criticism for her statements supporting Russian President Vladimir Putin as well as her 2017 meeting with Syria’s then-president, Bashar al-Assad. More alarming, the Post found evidence that Gabbard tried to obfuscate details about the nature of her encounters with the Syrian dictator from congressional investigators and may have lied to her staff. Having a history of shady meetings with any foreign national, much less the head of a country, is a great way not to be approved for a security clearance. (Just ask Trump’s son-in-law Jared Kushner, whose own opaque interactions with foreign officials temporarily stopped him from obtaining a clearance in the first Trump administration.)

During her confirmation hearing, Gabbard resisted entreaties from her fellow Republicans and Democrats—with whom she used to caucus when she was a member of Congress—to condemn Edward Snowden’s leaks and label him a “traitor.” Gabbard, who has long praised Snowden as a courageous whistleblower and called on Trump to pardon him, would say only that he “broke the law,” an obstinate position that left the distinct impression she approves of what Snowden did. Nevertheless, today the Senate voted largely along party lines to confirm Gabbard’s nomination as the nation’s top intelligence official.

Traditionally, counterintelligence officials have judged people whose ideology mirrors that of an adversarial state, or who have financial conflicts of interest, to be at higher risk of becoming spies or leaking secrets. “At the moment, that’s the population from which President Trump is selecting his most powerful and influential members of his administration,” Naftali told me.

[Read: It’s time to worry about DOGE’s AI plans]

Trump’s assault on the country’s national-security agencies stems from a distrust that millions of Americans share, Jeffrey Rogg, an intelligence historian at the University of South Florida, told me. Trump has repeatedly said—accurately—that the intelligence community often falls short of its basic obligation of keeping the United States from being taken by surprise by the country’s adversaries. And the agencies have failed several times to root out their own insider threats. Those counterintelligence debacles shake public confidence and bolster Trump’s critique that the intelligence agencies are dysfunctional and even corrupt.

At the same time, many career intelligence officers don’t trust the president or the people he has chosen to lead. They believe that Trump has misled the public about what the intelligence agencies are really there to do. And these, too, are accurate complaints, shared by many Americans.

Intelligence agencies depend on trust, both in their own employees and from the public. That confidence is disintegrating. As Rogg told me, “This is where we’re going to be our own worst enemies.”

Elon Musk’s unceasing attempts to access the data and information systems of the federal government range so widely, and are so unprecedented and unpredictable, that government computing experts believe the effort has spun out of control. This week, we spoke with four federal-government IT professionals—all experienced contractors and civil servants who have built, modified, or maintained the kind of technological infrastructure that Musk’s inexperienced employees at his newly created Department of Government Efficiency are attempting to access. In our conversations, each expert was unequivocal: They are terrified and struggling to articulate the scale of the crisis.

Even if the president of the United States, the head of the executive branch, supports (and, importantly, understands) these efforts by DOGE, these experts told us, they would still consider Musk’s campaign to be a reckless and dangerous breach of the complex systems that keep America running. Federal IT systems facilitate operations as varied as sending payments from the Treasury Department and making sure that airplanes stay in the air, the sources told us.

Based on what has been reported, DOGE representatives have obtained or requested access to certain systems at the U.S. Treasury, the Department of Health and Human Services, the Office of Personnel Management, and the National Oceanic and Atmospheric Administration, with eyes toward others, including the Federal Aviation Administration. “This is the largest data breach and the largest IT security breach in our country’s history—at least that’s publicly known,” one contractor who has worked on classified information-security systems at numerous government agencies told us this week. “You can’t un-ring this bell. Once these DOGE guys have access to these data systems, they can ostensibly do with it what they want.”

[Read: If DOGE goes nuclear]

What exactly they want is unclear. And much remains unknown about what, exactly, is happening here. The contractor emphasized that nobody yet knows which information DOGE has access to, or what it plans to do with it. Spokespeople for the White House, and Musk himself, did not respond to emailed requests for comment. Some reports have revealed the scope of DOGE’s incursions at individual agencies; still, it has been difficult to see the broader context of DOGE’s ambition.

The four experts laid out the implications of giving untrained individuals access to the technological infrastructure that controls the country. Their message is unambiguous: These are not systems you tamper with lightly. Musk and his crew could act deliberately to extract sensitive data, alter fundamental aspects of how these systems operate, or provide further access to unvetted actors. Or they may act with carelessness or incompetence, breaking the systems altogether. Given the scope of what these systems do, key government services might stop working properly, citizens could be harmed, and the damage might be difficult or impossible to undo. As one administrator for a federal agency with deep knowledge about the government’s IT operations told us, “I don’t think the public quite understands the level of danger.”

Each of our four sources, three of whom requested anonymity out of fear of reprisal, made three points very clear: These systems are immense, they are complex, and they are critical. A single program run by the FAA to help air-traffic controllers, En Route Automation Modernization, contains nearly 2 million lines of code; an average iPhone app, for comparison, has about 50,000. The Treasury Department disburses trillions of dollars in payments per year.

Many systems and databases in a given agency feed into others, but access to them is restricted. Employees, contractors, civil-service government workers, and political appointees have strict controls on what they can access and limited visibility into the system as a whole. This is by design, as even the most mundane government databases can contain highly sensitive personal information. A security-clearance database such as those used by the Department of Justice or the Bureau of Alcohol, Tobacco, Firearms and Explosives, one contractor told us, could include information about a person’s mental-health or sexual history, as well as disclosures about any information that a foreign government could use to blackmail them.

Even if DOGE has not tapped into these particular databases, The Washington Post reported on Wednesday that the group has accessed sensitive personnel data at OPM. Mother Jones also reported on Wednesday that an effort may be under way to effectively give Musk control over IT for the entire federal government, broadening his access to these agencies. Trump has said that Musk is acting only with his permission. “Elon can’t do and won’t do anything without our approval,” he said to reporters recently. “And we will give him the approval where appropriate. Where it’s not appropriate, we won’t.” The specter of what DOGE might do with that approval is still keeping the government employees we spoke with up at night. With relatively basic “read only” access, Musk’s people could easily find individuals in databases or clone entire servers and transfer that secure information somewhere else. Even if Musk eventually loses access to these systems—owing to a temporary court order such as the one approved yesterday, say—whatever data he siphons now could be his forever.

[Read: Trump advisers stopped Musk from hiring a noncitizen at DOGE]

With a higher level of access—“write access”—a motivated person may be able to put their own code into the system, potentially without any oversight. The possibilities here are staggering. One could alter the data these systems process, or they could change the way the software operates—without any of the testing that would normally accompany changes to a critical system. Still another level of access, administrator privileges, could grant the broad ability to control a system, including hiding evidence of other alterations. “They could change or manipulate treasury data directly in the database with no way for people to audit or capture it,” one contractor told us. “We’d have very little way to know it even happened.”

The specific levels of access that Musk and his team have remain unclear and likely vary between agencies. On Tuesday, the Treasury said that DOGE had been given “read only” access to the department’s federal payment system, though Wired then reported that one member of DOGE was able to write code on the system. Any focus on access tiers, for that matter, may actually simplify the problem at hand. These systems aren’t just complex at the code level—they are multifaceted in their architecture. Systems can have subsystems; each of these can have their own permission structures. It’s hard to talk about any agency’s tech infrastructure as monolithic. It’s less a database than it is a Russian nesting doll of databases, the experts said.

Musk’s efforts represent a dramatic shift in the way the government’s business has traditionally been conducted. Previously, security protocols were so strict that a contractor plugging a non-government-issued computer into an ethernet port in a government agency office was considered a major security violation. Contrast that with DOGE’s incursion. CNN reported yesterday that a 23-year-old former SpaceX intern without a background check was given a basic, low tier of access to Department of Energy IT systems, despite objections from department lawyers and information experts. “That these guys, who may not even have clearances, are just pulling up and plugging in their own servers is madness,” one source told us, referring to an allegation that DOGE had connected its own server at OPM. “It’s really hard to find good analogies for how big of a deal this is.” The simple fact that Musk loyalists are in the building with their own computers is the heart of the problem—and helps explain why activities ostensibly authorized by the president are widely viewed as a catastrophic data breach.

The four systems professionals we spoke with do not know what damage might already have been done. “The longer this goes on, the greater the risk of potential fatal compromise increases,” Scott Cory, a former CIO for an agency in the HHS, told us. At the Treasury, this could mean stopping payments to government organizations or outside contracts it doesn’t want to pay. It could also mean diverting funds to other recipients. Or gumming up the works in the attempt to do those, or other, things.

In the FAA, even a small systems disruption could cause mass grounding of flights, a halt in global shipping, or worse, downed planes. For instance, the agency oversees the Traffic Flow Management System, which calculates the overall demand for airspace in U.S. airports and which airlines depend on. “Going into these systems without an in-depth understanding of how they work both individually and interconnectedly is a recipe for disaster that will result in death and economic harm to our nation,” one FAA employee who has nearly a decade of experience with its system architecture told us. “‘Upgrading’ a system of which you know nothing about is a good way to break it, and breaking air travel is a worst-case scenario with consequences that will ripple out into all aspects of civilian life. It could easily get to a place where you can’t guarantee the safety of flights taking off and landing.” Nevertheless, on Wednesday Musk posted that “the DOGE team will aim to make rapid safety upgrades to the air traffic control system.”

Even if DOGE members are looking to modernize these systems, they may find themselves flummoxed. The government is big and old and complicated. One former official with experience in government IT systems, including at the Treasury, told us that old could mean that the systems were installed in 1962, 1992, or 2012. They might use a combination of software written in different programming languages: a little COBOL in the 1970s, a bit of Java in the 1990s. Knowledge about one system doesn’t give anyone—including Musk’s DOGE workers, some of whom were not even alive for Y2K—the ability to make intricate changes to another.

[Read: The “rapid unscheduled disassembly” of the United States government]

The internet economy, characterized by youth and disruption, favors inventing new systems and disposing of old ones. And the nation’s computer systems, like its roads and bridges, could certainly benefit from upgrades. But old computers don’t necessarily make for bad infrastructure, and government infrastructure isn’t always old anyway. The former Treasury official told us that mainframes—and COBOL, the ancient programming language they often run—are really good for what they do, such as batch processing for financial transactions.

Like the FAA employee, the payment-systems expert also fears that the most likely result of DOGE activity on federal systems will be breaking them, especially because of incompetence and lack of proper care. DOGE, he observed, may be prepared to view or hoover up data, but it doesn’t appear to be prepared to carry out savvy and effective alterations to how the system operates. This should perhaps be reassuring. “If you were going to organize a heist of the U.S. Treasury,” he said, “why in the world would you bring a handful of college students?” They would be useless. Your crew would need, at a minimum, a couple of guys with a decade or two of experience with COBOL, he said.

Unless, of course, you had the confidence that you could figure anything out, including a lumbering government system you don’t respect in the first place. That interpretation of DOGE’s theory of self seems both likely and even more scary, at the Treasury, the FAA, and beyond. Would they even know what to do after logging in to such a machine? we asked. “No, they’d have no idea,” the payment expert said. “The sanguine thing to think about is that the code in these systems and the process and functions they manage are unbelievably complicated,” Scott Cory said. “You’d have to be extremely knowledgeable if you were going into these systems and wanting to make changes with an impact on functionality.”

But DOGE workers could try anyway. Mainframe computers have a keyboard and display, unlike the cloud-computing servers in data centers. According to the former Treasury IT expert, someone who could get into the room and had credentials for the system could access it and, via the same machine or a networked one, probably also deploy software changes to it. It’s far more likely that they would break, rather than improve, a Treasury disbursement system in so doing, one source told us. “The volume of information they deal with [at the Treasury] is absolutely enormous, well beyond what anyone would deal with at SpaceX,” the source said. Even a small alteration to a part of the system that has to do with the distribution of funds could wreak havoc, preventing those funds from being distributed or distributing them wrongly, for example. “It’s like walking into a nuclear reactor and deciding to handle some plutonium.”

DOGE is many things—a dismantling of the federal government, a political project to flex power and punish perceived enemies—but it is also the logical end point of a strain of thought that’s become popular in Silicon Valley during the boom times of Big Tech and easy money: that building software and writing code aren’t just dominant skills for the 21st century, but proof of competence in any realm. In a post on X this week, John Shedletsky, a developer and an early employee at the popular gaming platform Roblox, summed up the philosophy nicely: “Silicon Valley built the modern world. Why shouldn’t we run it?”

This attitude disgusted one of the officials we spoke with. “There’s this bizarre belief that being able to do things with computers means you have to be super smart about everything else.” Silicon Valley may have built the computational part of the modern world, but the rest of that world—the money, the airplanes, the roads, and the waterways—still exists. Knowing something, even a lot, about computers guarantees no knowledge about the world beyond them.

“I’d like to think that this is all so massive and complex that they won’t succeed in whatever it is they’re trying to do,” one of the experts told us. “But I wouldn’t want to wager that outcome against their egos.”

Elon Musk is not the president, but it does appear that he—a foreign-born, unelected billionaire who was not confirmed by Congress—is exercising profound influence over the federal government of the United States, seizing control of information, payments systems, and personnel management. It is nothing short of an administrative coup.

As the head of an improvised team within the Trump administration with completely ambiguous power (the Department of Government Efficiency, or DOGE, in reference to a meme about a Shiba Inu), Musk has managed quite a lot in the two weeks since Inauguration Day. He has barged into at least one government building and made plans to end leases or sell some of them (three leases have been terminated so far, according to Stephen Ehikian, the General Services Administration’s acting administrator). He has called in employees from Tesla and the Boring Company to oversee broad workforce cuts, including at the Office of Personnel Management (one of Musk’s appointed advisers, according to Wired, is just 21 years old, while another graduated from high school last year). During this time, OPM staffers, presumably affiliated with DOGE, reportedly set up an “on-premise” email server that may be vulnerable to hacking and able to collect data on government employees—one that a lawsuit brought by two federal workers argues violates the E-Government Act of 2002 (there has not yet been a response to the complaint). Musk’s people have also reportedly gained access to the Treasury’s payments system—used to disburse more than $5 trillion to Americans each year (a national-security risk, according to Senator Ron Wyden, a democrat from Oregon)—as well as computer systems that contain the personal data of millions of civil servants. (They subsequently locked some senior employees out of those systems, according to Reuters.) Musk did not immediately respond to a request for comment.

Over the weekend, the Trump administration put two senior staffers at the U.S. Agency for International Development (USAID) on administrative leave—staffers who, according to CNN, had tried to thwart Musk’s staff’s attempts to access sensitive and classified information. Musk posted on X yesterday that “USAID is a criminal organization. Time for it to die.” USAID staffers were barred from entering the unit’s headquarters today.

This is called “flooding the zone.” Taken in aggregate, these actions are overwhelming. But Musk’s political project with DOGE is actually quite straightforward: The world’s richest man appears to be indiscriminately dismantling the government with an eye toward consolidating power and punishing his political enemies.

Two days before the 2024 election, I wrote that Musk’s chaotic takeover of Twitter was going to be the blueprint for his potential tenure at DOGE. Unfortunately, I was right—he’s running the exact same playbook. But it’s worth keeping in mind that there are two ways of measuring success for Musk’s projects: first, whether the organizations themselves benefit under his leadership, and second, whether Musk himself gets something out of the arrangement. Musk’s stewardship of X has been a financial nightmare. He has alienated advertisers, tanked revenue and user growth, and saddled investment banks with debt from the purchase that they’ll need to sell off. Yet Musk’s own influence and net worth have grown considerably during this time. His fanboys and the MAGA faithful don’t care that X is a flailing business, because Musk did deliver on giving liberals their supposed comeuppance by de-verifying accounts and reinstating banned trolls. He turned the platform into a conspiratorial superfund site, has boosted right-wing accounts and talking points, and helped elect Donald Trump as president. Musk’s purchase is a success in their eyes because he succeeded in turning X into a political weapon.

[Read: Musk’s Twitter is the blueprint for a MAGA government]

The same thing is happening right now with DOGE. Musk and his Silicon Valley acolytes are acting on a long-held fantasy of approaching the federal government like a software company and running it like a venture-backed tech start-up during the days of zero-percent interest rates. Here’s the problem: The federal government is not a software company. “The stakes are wildly different,” a former senior Twitter executive told me recently. This person, who requested anonymity because they worked closely with Musk during his takeover and fear retribution, argued that Musk seems incapable of recognizing the limits of his own knowledge. When I asked them to describe Musk’s managerial strategy, they borrowed a term of art from SpaceX’s own rocket mishaps: “This is a rapid unscheduled disassembly of government services.”

The theory that the government is inefficient is not altogether incorrect. I recently spoke with Robert Gordon, formerly the deputy assistant to the president for economic mobility in the Biden administration, to get a sense of how intricate government agencies are and what it would take to reform them. Gordon, who has spent time in the Office of Management and Budget and as the assistant secretary responsible for grants policy at the Department of Health and Human Services, was quick to note that we desperately need to simplify processes within the federal government to allow workers to execute more quickly and develop more agile technology, such as the Direct File product that the IRS recently made to allow Americans to file taxes for free. “No doubt the government could do more here,” he told me. “But it requires incredibly specific approaches, implemented in a thoughtful way. It requires paying enormous attention to detail, not blowing shit up.” Musk and DOGE have instead operated with a “vast carelessness,” Gordon wrote in a Substack post last week. “This government cannot trouble itself to plan for the biggest things, the funds that thousands of organizations use to serve millions of people,” he wrote. “It has swept up civil servants in a vortex of confusion and fear.” Musk wrote today on X that the Treasury team that built Direct File no longer exists. “That group has been deleted,” he said.

[Read: The American people deserve DOGE]

Among Gordon’s biggest concerns is that DOGE’s slapdash cuts will remove key links in the bureaucratic chain that make the government function. Even simple-sounding procedures—allocating government funds in a crisis like, say, a pandemic—require coordination among teams of civil servants across multiple government offices. “All of this is done by back-office types,” Gordon told me. “There are so many people in that process, and it matters enormously how good they are.” That this system is inefficient is frustrating, Gordon said, but he worries that the chaos caused by Musk’s efforts will halt any possibility of reform. “If you want to make this system better, you need to create space for civil servants who know what they’re doing to do that work,” he told me. “What’s very likely to happen now because of this pressure is that the most competent people on that chain are at super-high risk of saying, I gave it my best shot; I don’t need this and quit, because they can get better jobs. That’s what I see happening.”

Of course, the so-called tech right does not agree. As the political scientist Henry Farrell wrote this past weekend, “The fact that none of the DOGE people actually understand how government functions is a feature, not a bug. If you understand the workings of the federal bureaucracy, you are almost certainly part of the problem, not the solution.” But this reasoning is not usually compatible with the reality of managing complex organizations. As the former Twitter exec told me, after Musk took over the platform, his people enthusiastically championed ideas that seasoned employees with knowledge of the company had already researched and rejected: “It wasn’t that we hadn’t thought about new ways, say, to do verification or handle bots, but we rejected them on the basis of research and data. There was a huge contrast between the methodical approach and Musk’s rapid-fire whims.”

When Musk barged into Twitter in 2022 as its new CEO, his strategy was “decision making by vibes,” according to the former exec I spoke with. Those vibes were often dictated by the sycophants in Musk’s orbit. The executive described Musk as surprisingly receptive to ideas when presented with facts and data, but said that few in his inner circle questioned or spoke frankly with him: “And so, in the absence of rational decision making, we got the vibes-based, yes-man approach.”

The former executive did point to a meaningful difference between X and DOGE, however: The government is big and complex. This may be an asset during an assault. “Even if you try to take a flamethrower to the government, the destruction won’t be quick. There’ll be legal challenges and congressional fights, and in the months and weeks, it’ll be individuals who keep essential services running,” they said. The government workers who know what they’re doing may still be able to make positive incremental change from within.

[Read: There really is a deep state]

It’s a rousing, hopeful notion. But I fear that the focus on the particulars of this unqualified assault on our government is like looking at X’s bottom line, in that it obscures Musk’s real ambitions. What are DOGE’s metrics for success? If X is our guide, health, functionality, and sustainability are incidental and able to be sacrificed. The end game for Musk seems to be just as it was with Twitter: seize a polarized, inefficient institution; fuse his identity with it; and then use it to punish his enemies and reward his friends. DOGE is a moon-shot program to turn the government into Musk’s personal political weapon.

This is an edition of The Atlantic Daily, a newsletter that guides you through the biggest stories of the day, helps you discover new ideas, and recommends the best in culture. Sign up for it here.

Senator Katie Britt, Republican of Alabama, is upset. She believes that someone in the United States Air Force decided to interpret President Donald Trump’s recent executive order to terminate “all discriminatory programs, including illegal DEI and ‘diversity, equity, inclusion, and accessibility’ (DEIA) mandates, policies, programs, preferences, and activities in the Federal Government, under whatever name they appear,” just the way it was written.

No one is quite sure what happened, but somehow this order resulted in the excision from a U.S. Air Force training course of some materials about the legendary Tuskegee Airmen, the all-Black World War II fighter pilots known as the Red Tails because of their aircraft’s distinctive markings. Air Force officials confirmed on Saturday that a video had been removed from the training curriculum but only because it was “intertwined in courses now under review,” and it is now back in the curriculum.

Britt referred to this kind of action as “malicious compliance,” meaning a kind of opposition through aggressive and sometimes overly literal implementation of a command or policy. Rather than refuse to obey, the person or group engaging in malicious compliance takes a kind of “monkey’s paw” approach, implementing the directives as destructively as possible. (Every teenager who has loaded the dishwasher improperly on purpose, hoping never to be told to clear the table again, knows what malicious compliance means.)

Britt also tagged Secretary of Defense Pete Hegseth on social media. Hegseth, who was nominated for his position in part because of his vow to root out wokeness and DEI and to replace them with “lethality,” responded enthusiastically: “Amen! We’re all over it Senator. This will not stand.”

Britt’s complaint about malicious compliance is a diversion. Trump’s wave of executive orders is designed to be performatively malicious. My colleague Adam Serwer years ago noted that, for the MAGA movement, “the cruelty is the point,” and now Trump’s orders make clear that the malice is the policy.

The series of presidential decrees is largely intended to delight the Republican base; unfortunately, government workers cannot divine what Trump really meant. The president has not given any cue that his orders should be interpreted in some more generous way. In fact, days before the Air Force kerfuffle, federal workers received an email from their supervisors (based on a template provided by the Office of Personnel Management) that could have come straight from a party apparatchik in the old Soviet Union. This memo not only told staff to be on the lookout for attempts to hide DEI-related ideological contamination, but warned them of their obligation to rat out colleagues who did so or face “adverse” job consequences themselves.

The advisory, which has since been taken off a government website, continued: “We are aware of efforts by some in government to disguise these programs by using coded or imprecise language. If you are aware of a change in any contract description or personnel position description since November 5, 2024”—that is, since Election Day—“to obscure the connection between the contract and DEIA or similar ideologies,” employees must report it to OPM within ten days.

This is not exactly language that encourages anyone to use common sense and good judgment to decide what constitutes DEI contraband. This is a command that says, in effect: This could mean anything; if you don’t report it, and we find it, you’re in trouble. When government employees get a memo like that, they are not inclined to sit around wondering what counts and what doesn’t.

Trump’s other executive orders are likewise designed to show the GOP base that the new administration is doing all of the things that Trump promised he’d do—even if they’re things that, legally, no president can do. Trump had pledged, for example, to eliminate birthright citizenship, so he sharpied out part of the Fourteenth Amendment and declared victory. He froze federal grants and loans—an order now temporarily blocked by a judge—which could have endangered any number of programs, including school lunches. (And about time, according to Representative Rich McCormick, Republican of Georgia, who told CNN today that those indolent kids need to go get jobs—even, apparently, schoolchildren who aren’t old enough to work—instead of “spong[ing] off the government”).

What would non-malicious compliance with such a mandate even look like? Instead of a lunch, are schools supposed to hand poor kids a glass of water and then wish them luck in their job search?

Of course, the Trump administration knows that aid to states and localities will begin to flow again, that children will be getting lunches, and that babies born on U.S. soil are citizens. The goal of all these orders is not to implement policy, but to generate outrage, report the spasms of liberal apoplexy to the MAGA faithful, and then, when necessary, go to court. And why not? The president now has a politically sympathetic Supreme Court majority that worked hard to keep him out of prison while he was a candidate, and has functionally immunized him against almost any challenge now that he’s back in office. Trump’s people know that they cannot actually shake the Constitution like an Etch A Sketch and make birthright citizenship disappear, but why not give it a shot, especially if a trolling executive order makes the base happy?

Trump and his people may also believe that a sleet storm of executive orders, some of which might stick here and there while others melt on contact with reality, is a way to demonstrate competence. They are likely still stung by the fiasco over the 2017 travel ban that initially got swatted down in court, and this time they want to appear as if they know what they’re doing.

But this is merely mimicking competence and energy. The “return to work” order, for example, is a MAGA fan favorite, because it plays to a common stereotype among many Americans that federal employees who work from home are scamming goldbrickers plodding around the house in their bunny slippers and tapping the occasional key on a laptop. Although showing up to an office or worksite in-person is (and should be) a basic requirement of most jobs, remote work in many cases benefits the government and the taxpayer: It reduces congestion in cities, and it offloads a lot of overhead costs (heat, water, lighting, etc.) onto the worker. That’s why the government and private industry were trending toward remote arrangements long before the pandemic.

In any case, many federal offices don’t have enough space to bring everyone back, but Trump may be attempting to make government service onerous enough that some of them will leave anyway: All federal employees have until February 6 to accept a sizable buyout if they cannot or will not return to in-person work. In the end, the RTO power play isn’t really about trying to fill empty offices. Instead, Trump is telling federal employees that all of the arrangements they’ve made with their departments about schedules, child care, commutes, and staffing are now invalid, because their career and service matters less than making some red-state voter feel that the president finally stuck it to them and their co-workers.

Maybe a non-malicious way to enforce such orders exists. But that’s not the point.

Related:

The cruelty is the point. (From 2018) The strategy behind Trump’s policy blitz

Here are three new stories from The Atlantic:

Trump tries to seize the “power of the purse.” What an undervaccinated America would look like China’s DeepSeek surprise

Today’s News

A district-court judge temporarily blocked the Trump administration’s pause on federal grants and loans. Trump signed an executive order that would exclude gender-transition care from federal insurance programs. The Department of Justice announced yesterday that it has fired more than a dozen officials who worked on the criminal investigations into Trump.

Evening Read

Illustration by Danielle Del Plato

Just Say No to Terrible White LEDs

By Gilad Edelman

God said, “Let there be light”—everyone knows that. But God did not specify what color light, and this would eventually prove problematic.

In the age of the LED light bulb, consumers have an unfathomable range of lighting options. This has, perversely, made the task of pleasantly illuminating our homes harder, not easier.

Read the full article.

More From The Atlantic

“Dear James”: Oh, how the men drone on The libs are having their paranoia moment. Blind partisanship does not actually help Trump. Biden’s Middle East legacy

Culture Break

Jordan Hemingway

Listen. The sensation you get when dancing or making a really good cup of tea? FKA Twigs wants to bottle that in Eusexua, her latest album.

Debate. The Oscars have left the mainstream moviegoer behind, David Sims argues.

Play our daily crossword.

Stephanie Bai contributed to this newsletter.

Explore all of our newsletters here.

When you buy a book using a link in this newsletter, we receive a commission. Thank you for supporting The Atlantic.