For years, Donald Trump’s critics have accused him of behaving like a crooked used-car salesman. Yesterday afternoon, he did it for real on the White House South Lawn.

Squinting in the sun with Elon Musk, Trump stood next to five Tesla vehicles, holding a piece of paper with handwritten notes about their features and costs. Trump said he would purchase a car himself at full price. Then Trump and Musk got into one of the cars. Musk explained that the electric vehicle was “like a golf cart that goes really fast.” Trump offered his own praise to the camera: “Wow. That’s beautiful. This is a different panel than I’ve—everything’s computer!”

This was a stilted, corrupt attempt to juice a friend’s stock, and certainly beneath the office of the presidency. But you ought not to overlook just how embarrassing the spectacle was for Musk. The subtext of the event—during which Trump also declared that the White House would label any acts of violence against Tesla dealerships as domestic terrorism—was the ongoing countrywide protests against Tesla, due to Musk’s role in the Trump administration. In some cities, protesters have defaced or damaged Tesla vehicles and set fire to the company’s charging stations. Tesla’s stock price has fallen sharply—almost 50 percent since its mid-December, postelection peak—on the back of terrible sales numbers in Europe. The hastily assembled White House press event was presented as a show of solidarity, but the optics were quite clear: Musk needed Trump to come in and fix his mess for him.

And Tesla isn’t the only Musk venture that’s struggling. SpaceX’s massive new Starship rocket has exploded twice this year during test flights. And Ontario, Canada, has canceled its contract with his Starlink internet company to provide service to remote communities, citing Trump’s tariffs. According to the Bloomberg Billionaire Index, Musk is $148 billion poorer than he was on Inauguration Day (he is currently worth $333.1 billion).

Just 17 days after wielding a chain saw and dancing triumphantly onstage at CPAC, the billionaire looked like he was about to cry on the Fox Business channel earlier this week. He confessed that he was having “great difficulty” running his many businesses, and let out a long, dismal sigh and shrugged when asked if he might go back to his businesses after he’s done in the administration.

The world’s richest man can be cringe, stilted, and manic in public appearances, but rarely have I seen him appear as defeated as he has of late, not two months into his role as a presidential adviser. In the past few weeks, he’s been chastised by some of Trump’s agency heads for overstepping his bounds as an adviser (Trump sided with the agency heads). Reports suggest that some Republican lawmakers are frustrated with Musk’s bluster and that the DOGE approach to slashing the federal bureaucracy is angering constituents and making lawmakers less popular in their districts. DOGE has produced few concrete “wins” for the Trump administration and has instead alienated many Americans who see Musk as presiding over a cruel operation that is haphazardly firing and rehiring people and taking away benefits. Numerous national polls in recent weeks indicate that a majority of respondents disapprove of Musk’s role and actions in the government.

Musk’s deep sighs on cable TV and emergency Tesla junkets on the White House lawn are hints that he may be beginning to understand the precariousness of his situation. He is well known for his high risk tolerance, overleveraging, and seemingly wild business bets. But his role at DOGE represents the biggest reputational and, consequently, financial gamble of his career. Musk is playing a dangerous game, and he looks to be losing control of the narrative.

And the narrative is everything. Elon Musk is many things—the richest man in the world, an internet-addled conspiracy theorist, the controller of six companies, perhaps even the shadow president of the United States—but most importantly, he is an idea. The value of Musk may be tied more to his image than his actual performance. He’s a human meme stock.

A CNN clip from late October captures this notion. In it, a reporter is standing outside one of Musk’s America PAC rallies in Pennsylvania, interviewing the CEO’s superfans, most of whom are unequivocal that Musk is “the smartest man in the world.” He has an engineer’s mindset, one attendee claims, meaning he sees the world differently. Two other men in the clip say that Musk got them to pay attention to politics (and to Trump, specifically). These people had fallen hard for a cultivated image of  Musk as a Thomas Edison or Tony Stark type, a great man of history who is single-handedly pushing the bounds of progress. Musk has had great success popularizing electric vehicles and building new rockets (though many still debate his direct involvement in the engineering). These supporters might have been fans of his companies, but they seem to have also fallen for the myth of his genius, a story born out of years of hagiographic books, news articles reporting his hyperbolic claims, and Musk’s own ability to command attention.

[Read: Elon Musk’s texts shatter the myth of the tech genius]

The image of Musk as a true visionary has proved surprisingly durable. In the early to mid-2010s, Musk took advantage of a different era of technology coverage—one that was more gadget-focused and largely uncritical—to hype his ideas for the future of transportation and interplanetary exploration. At that time, Tesla, his signature project, was coded as progressive and marketed as being in line with climate-change goals. The cultural dynamic of these ideas has changed, but the fundamental product being sold by Musk has not: one man with a singular ability to brute force his way to the future.

Musk’s trajectory changed after Trump was first elected president. It was during this period that Musk—already an incessant poster—realized how Twitter could be used to command an unbelievable amount of attention. Even when that attention was negative, the process of repeatedly making himself the main character on the platform elevated Musk’s profile. He became more polarizing (for chastising journalists, behaving erratically, making a supposed weed joke on Twitter that got him in trouble with the SEC, and getting sued for defamation for calling somebody a “pedo guy”), yet this somehow only added to Musk’s lore.

For years, valid criticisms of the Tesla executive came with an asterisk: He’s erratic, crude, even a little unstable, but that’s all part of the larger visionary package. Kara Swisher showcased this dynamic well in a 2018 New York Times column titled “Elon Musk Is the Id of Tech.” “I find the hagiography around him tiresome and even toxic,” she wrote. But also, “Mr. Musk’s mind and ideas are big ones.” As Swisher noted, Musk’s attention-seeking at the time had a secondary effect of alienating him from some of his peers and fans. But tweet by tweet, Musk found a different audience, one eager to embrace his visionary image, provided he took up their crusade against “wokeism.”

During the pandemic, Musk’s posting frequency intensified considerably as he began to stake out more reactionary territory. He called the COVID-19 panic in March 2020 “dumb” and later that year tweeted that “pronouns suck.” Musk endeared himself to the right wing by positioning himself as a free-speech warrior, a posture that ultimately led him to purchase Twitter. Right-wing influencers and the MAGA faithful saw Musk’s turn as proof of their movement’s ascendance, but what has happened since Musk turned Twitter into X is nothing short of audience capture: Musk has fully become the person his right-wing fanboys want him to be, pushing far beyond a mere  dalliance with conspiracy theories and “Great Replacement” rhetoric. It is hardly controversial to suggest based on Musk's posts and blatant political activism that the centibillionaire has been further radicalized by his platform, which he then turned into a political weapon to help elect Trump.

Musk’s X and MAGA bets mostly paid off, at least in the near term. Before Musk bought Twitter, I highlighted a comment from Lily Francus, then the director of quant research at Moody’s Analytics, who noted, “I do think fundamentally that a significant fraction of Tesla’s value is due to the fact that Elon can command this attention continuously.” Francus doesn’t go as far as to say that Tesla behaves like a meme stock—which can surge in price after going viral as a result of coordinated efforts online—but that Musk himself has this quality. Musk’s Twitter purchase was a bad deal financially and has been detrimental to X’s bottom line, but his ownership of the platform helped boost his cultural and political relevance by keeping him in the center of the news cycle. Similarly, Musk going all in on Donald Trump, becoming a megadonor to Republicans, and ultimately getting the DOGE gig all resulted in Tesla stock soaring—up until a point.

You can argue that there’s a flywheel effect to all of this. Musk’s polarizing, upsetting, attention-seeking behavior has made him unavoidable and increased his political influence, which, in turn, has increased his net worth overall. This has only improved Musk’s standing with Trump, who both respects great wealth and appears flattered by the notion that the richest man in the world wants to spend his time shadowing him around Washington and Mar-a-Lago.

Musk is used to being leveraged, trading on his reputation or his illiquid assets to keep the flywheel spinning. To his credit, he tends to make it work. He’s flouted the law when that has been advantageous to his business interests and taken advantage of a culture of elite impunity. He’s long been unafraid to get sued or reprimanded by a government agency. But two important things are different in his current situation. The first is the stakes of his reputational bet—rather than alienating himself from progressives or the media, Musk is threatening to meddle with essential government services, such as Social Security, that millions of Americans rely on. Indeed, Musk floated the idea of cutting Social Security benefits in his Fox Business interview on Monday. Whether he’s in charge of cuts or not, as DOGE’s figurehead, Musk risks infuriating countless people who object to the federal firings. Breaking the government is orders of magnitude different than buying a niche but influential microblogging platform.

[Read: There are no more red lines]

The second difference is the man he’s tied his reputation to: Trump. Musk’s attention-seeking and fondness for organizational chaos are usually unmatched, giving him an advantage in most of his dealings. This is not the case with Trump, whose shamelessness and penchant for discarding close confidants when they become liabilities are well documented. Musk is rich and powerful, but he is not the durable, singular political figure that Trump is. It is not difficult to imagine a scenario where this ends poorly for Musk. The flywheel could reverse: Musk could become universally reviled, causing the protests to increase and his net worth to shrink. The richest man in the world is valuable to Trump, as is the myth of Musk as the modern Edison. But mere billionaires? They are fungible tokens and easily interchangeable dais members in Trump’s eyes—just ask Mark Zuckerberg.

It would be foolish to suggest with any certainty that Musk is cooked. Historically, he’s managed to wriggle out of trouble. Perhaps the most hopeful outcome for Musk is that Trump has too much of his own presidency tied to Musk to throw him under the bus. It’s too early to say.

Yesterday’s White House stunt had all the hallmarks of Trump corruption, but there was something else, too—an air of desperation. It was a tacit admission that the protests are working and that Musk and Trump are rattled enough by current sentiment that they’re willing to turn the South Lawn into a showroom. Watching Musk clam up on Fox Business or quietly idle next to Trump in front of the White House, it’s even easier than normal to see past Musk’s trademark bullshitting and bluster. These moments make clear that this time, Musk has wagered the only thing he can’t easily buy back—the very myth he created for himself.

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.

As the head of the Office of Special Counsel, Hampton Dellinger had a triple target on his back from the start of Donald Trump’s presidency: He was a Joe Biden appointee, he was the head of one of the independent regulatory agencies that the Trump administration is targeting, and his duty was to fight to protect the jobs of tens of thousands of civil servants the president has tried to fire.

So when Dellinger received an email on Friday, February 7, telling him that he’d been dismissed, he wasn’t surprised. He also wasn’t going to quietly concede. Under a law that’s stood for decades, the special counsel serves a five-year term and “may be removed by the President only for inefficiency, neglect of duty, or malfeasance in office.” The following Monday morning, Dellinger filed a suit challenging his firing, and by that night, a federal judge had temporarily reinstated him.

During the following month, Dellinger led a bifurcated life that he joked was “like a Severance episode, except I was always at work”: one workplace “where I was advocating for others, and that was the place I wanted to be completely focused,” he told me on Friday. “But then the other side of it was trying to keep my job.”

OSC is a classic post-Watergate creation, designed to insulate the functioning of the federal government from political and other improper interference. It’s charged with protecting whistleblowers inside the executive branch and with identifying violations of the Hatch Act, which prohibits politicking by government officials. If OSC believes that federal employees have been improperly fired, it can file a case with the Merit Systems Protection Board.

This makes an otherwise obscure office very important right now, because the Trump administration, with Elon Musk’s U.S. DOGE Service leading the charge, has laid off huge swaths of the federal workforce in apparent defiance of laws designed to protect them, with more cuts promised. Last Wednesday, Dellinger won a major victory: The MSPB ruled that the U.S. Department of Agriculture must temporarily rehire nearly 6,000 probationary employees while an investigation proceeds into whether they were wrongfully fired. He told me that he was ready to try to get tens of thousands more probationary employees reinstated.

Instead, Dellinger found himself out of a job a few hours later. On Wednesday night, a panel of judges on the D.C. Circuit Court of Appeals ruled for the administration and against Dellinger, declaring that he would be removed while he pursued his appeal. The next day, Dellinger announced that he was ending his fight.

“I knew it would take at least a year to get a final decision” in court, he told me. “It may well have gone against me, and by that point, seeing the damage that’s taking place on a daily basis at federal agencies, I knew there would be almost nothing I could do should I ever get back into my job.”

In his statement ending his challenge, he wrote: “I strongly disagree with the circuit court’s decision, but I accept and will abide by it. That’s what Americans do.” That was a pointed response to comments by several government officials, including Musk and Vice President J. D. Vance, who have questioned whether the executive branch has to follow judicial rulings. “I think the key to our country is respect for the rule of law, and I think there’s been too much disrespect of late,” he told me. “So I wanted to make it clear that just because I’m unhappy with the decision, I in no way contest its binding nature.”

What is at stake right now is not just the fate of whistleblowers and probationary employees but also the underlying principle of independent agencies within the executive branch. Such bodies have existed since the 1930s and are written into laws passed by Congress, but as I wrote recently, Trump allies have argued in Project 2025 and elsewhere that independent regulatory agencies are unconstitutional because they limit the president’s control of the executive branch. They have promised to politicize traditionally detached parts of the government.

If courts conclude that this independence is unconstitutional, then most existing protections for whistleblowing seem doomed. Congress concluded when passing these laws that the executive branch needed internal watchdogs. They are generally presidentially appointed—like Dellinger, and like inspectors general inside major departments—but, once in place, insulated from pressure. Without them, whistleblowers have no clear recourse besides going to Congress (no easy feat for all but the most major scandals) or the press. Either path is uncertain and fraught with dangers of retaliation.

Gutting the current regime may result in more of the problems that Musk is supposedly fighting, Dellinger argued. “I think it’ll mean that government is less effective,” he told me, because fewer routes will exist for employees to shed light on failures. “I think it may lead to an increase in waste, fraud, and abuse. And I think we’re not going to know for sure what it means, because you don’t have these independent watchdogs who are able to make their work public.”

The entire existing vision of the executive branch, constructed by an idealistic liberal vision of government held accountable by legal structures and processes, seems currently under threat. Dellinger is a fitting figure to be in the middle of this fight. He’s spent his career moving between government service and practicing law in the private sector. (He’s also contributed to The Atlantic.) His father, Walter Dellinger, served as the acting solicitor general in the Clinton administration and was regarded as one of the most brilliant Democratic lawyers of his generation. Hampton Dellinger told me he remains hopeful that the decades-old vision of the federal government is not dying.

“The fact that people are resisting unlawful orders, I think, is vital,” he said. “I still have faith in the judiciary, even if my case didn’t succeed. I have faith in generations younger than me.” If the federal government is to run on anything other than patrimonialism, those generations will have to find a way to rebuild it after the current assault.

Related:

Trump tests the courts. “It feels like it’s chaotic on purpose.”

Here are three new stories from The Atlantic.

The diseases are coming. Colleges have no idea how to comply with Trump’s orders. Trump drops the mask.

Today’s News

The U.S. stock market plunged today amid concerns over the economic pain that President Donald Trump’s aggressive tariff policies could bring.

ICE agents arrested the Columbia University graduate student and pro-Palestinian activist Mahmoud Khalil over the weekend. According to Khalil’s lawyer, agents said that they were operating under State Department orders to revoke his green card. The State Department declined to comment on Khalil’s case.

Elon Musk blamed a “massive cyberattack” for a series of outages on X.

More From The Atlantic

DOGE’s plans to replace humans with AI are already under way. Europe can’t trust the U.S. anymore. Teens are forgoing a classic rite of passage. Kara Swisher: Move fast and destroy democracy.

Evening Read

Lila Barth for The Atlantic

Turtleboy Will Not Be Stopped

By Chris Heath

On overpasses and by roadsides they gather, holding banners and placards. In the early days, only a few people showed up, congregating at chosen times and scattered locations around Boston. But their cause has grown and their numbers have swelled. For Labor Day 2024, plans were made for “standouts,” as the organizers called them, in more than 70 places—all over Massachusetts, yes, but also in Ohio, Kansas, Florida, California, and elsewhere.

These assemblies are the most visible manifestation of what is usually referred to as the Free Karen Read movement. If in the fullness of time it will seem strange that such unity and passion should have been mustered in defense of a 45-year-old Massachusetts financial analyst and adjunct college professor accused of killing her police-officer boyfriend by backing into him with her car … well, not to these people gathered today.

Read the full article.

Culture Break

Theo Wargo / Peacock / Getty

Listen. Lady Gaga’s latest album, Mayhem, is “an ode to her early career—and a powerful demonstration of growth,” our music critic Spencer Kornhaber writes.

Read. Waste Wars, by the journalist Alexander Clapp, tracks how the garbage of rich countries ends up in some of the world’s poorest places.

Play our daily crossword.

P.S.

This weekend, I drove to the North Carolina mountains with my family, and we spent most of the drive both ways listening to Big Ugly, the brand-new record by Fust, one of my favorite musical discoveries of the past year. The Durham-based alt-country band is led by Aaron Dowdy, who is a Ph.D. student in Duke University’s literature department but also firmly rooted in his native Appalachian Virginia. The lead track, “Spangled,” rhymes Route 11 with repossession and includes the memorable image of “feeling like a sparkler / that’s been thrown off a roof.” I’m obsessed.

— David

Isabel Fattal and Shan Wang contributed to this newsletter.

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

If you have tips about the remaking of the federal government, you can contact Matteo Wong on Signal at @matteowong.52.

A new phase of the president and the Department of Government Efficiency’s attempts to downsize and remake the civil service is under way. The idea is simple: use generative AI to automate work that was previously done by people.

The Trump administration is testing a new chatbot with 1,500 federal employees at the General Services Administration and may release it to the entire agency as soon as this Friday—meaning it could be used by more than 10,000 workers who are responsible for more than $100 billion in contracts and services. This article is based in part on conversations with several current and former GSA employees with knowledge of the technology, all of whom requested anonymity to speak about confidential information; it is also based on internal GSA documents that I reviewed, as well as the software’s code base, which is visible on GitHub.

[Read: DOGE has ‘god mode’ access to government data]

The bot, which GSA leadership is framing as a productivity booster for federal workers, is part of a broader playbook from DOGE and its allies. Speaking about GSA’s broader plans, Thomas Shedd, a former Tesla engineer who was recently installed as the director of the Technology Transformation Services (TTS), GSA’s IT division, said at an all-hands meeting last month that the agency is pushing for an “AI-first strategy.” In the meeting, a recording of which I obtained, Shedd said that “as we decrease [the] overall size of the federal government, as you all know, there’s still a ton of programs that need to exist, which is a huge opportunity for technology and automation to come in full force.” He suggested that “coding agents” could be provided across the government—a reference to AI programs that can write and possibly deploy code in place of a human. Moreover, Shedd said, AI could “run analysis on contracts,” and software could be used to “automate” GSA’s “finance functions.”

A small technology team within GSA called 10x started developing the program during President Joe Biden’s term, and initially envisioned it not as a productivity tool but as an AI testing ground: a place to experiment with AI models for federal uses, similar to how private companies create internal bespoke AI tools. But DOGE allies have pushed to accelerate the tool’s development and deploy it as a work chatbot amid mass layoffs (tens of thousands of federal workers have resigned or been terminated since Elon Musk began his assault on the government). The chatbot’s rollout was first noted by Wired, but further details about its wider launch and the software’s previous development had not been reported prior to this story.

The program—which was briefly called “GSAi” and is now known internally as “GSA Chat” or simply “chat”—was described as a tool to draft emails, write code, “and much more!” in an email sent by Zach Whitman, GSA’s chief AI officer, to some of the software’s early users. An internal guide for federal employees notes that the GSA chatbot “will help you work more effectively and efficiently.” The bot’s interface, which I have seen, looks and acts similar to that of ChatGPT or any similar program: Users type into a prompt box, and the program responds. GSA intends to eventually roll the AI out to other government agencies, potentially under the name “AI.gov.” The system currently allows users to select from models licensed from Meta and Anthropic, and although agency staff currently can’t upload documents to the chatbot, they likely will be permitted to in the future, according to a GSA employee with knowledge of the project and the chatbot’s code repository. The program could conceivably be used to plan large-scale government projects, inform reductions in force, or query centralized repositories of federal data, the GSA worker told me.

Spokespeople for DOGE did not respond to my requests for comment, and the White House press office directed me to GSA. In response to a detailed list of questions, Will Powell, the acting press secretary for GSA, wrote in an emailed statement that “GSA is currently undertaking a review of its available IT resources, to ensure our staff can perform their mission in support of American taxpayers,” and that the agency is “conducting comprehensive testing to verify the effectiveness and reliability of all tools available to our workforce.”

At this point, it’s common to use AI for work, and GSA’s chatbot may not have a dramatic effect on the government’s operations. But it is just one small example of a much larger effort as DOGE continues to decimate the civil service. At the Department of Education, DOGE advisers have reportedly fed sensitive data on agency spending into AI programs to identify places to cut. DOGE reportedly intends to use AI to help determine whether employees across the government should keep their job. In another TTS meeting late last week—a recording of which I reviewed—Shedd said he expects that the division will be “at least 50 percent smaller” within weeks. (TTS houses the team that built GSA Chat.) And arguably more controversial possibilities for AI loom on the horizon: For instance, the State Department plans to use the technology to help review the social-media posts of tens of thousands of student-visa holders so that the department may revoke visas held by students who appear to support designated terror groups, according to Axios.

Rushing into a generative-AI rollout carries well-established risks. AI models exhibit all manner of biases, struggle with factual accuracy, are expensive, and have opaque inner workings; a lot can and does go wrong even when more responsible approaches to the technology are taken. GSA seemed aware of this reality when it initially started work on its chatbot last summer. It was then that 10x, the small technology team within GSA, began developing what was known as the “10x AI Sandbox.” Far from a general-purpose chatbot, the sandbox was envisioned as a secure, cost-effective environment for federal employees to explore how AI might be able to assist their work, according to the program’s code base on GitHub—for instance, by testing prompts and designing custom models. “The principle behind this thing is to show you not that AI is great for everything, to try to encourage you to stick AI into every product you might be ideating around,” a 10x engineer said in an early demo video for the sandbox, “but rather to provide a simple way to interact with these tools and to quickly prototype.”

[Kara Swisher: Move fast and destroy democracy]

But Donald Trump appointees pushed to quickly release the software as a chat assistant, seemingly without much regard for which applications of the technology may be feasible. AI could be a useful assistant for federal employees in specific ways, as GSA’s chatbot has been framed, but given the technology’s propensity to make up legal precedents, it also very well could not. As a recently departed GSA employee told me, “They want to cull contract data into AI to analyze it for potential fraud, which is a great goal. And also, if we could do that, we’d be doing it already.” Using AI creates “a very high risk of flagging false positives,” the employee said, “and I don’t see anything being considered to serve as a check against that.” A help page for early users of the GSA chat tool notes concerns including “hallucination”—an industry term for AI confidently presenting false information as true—“biased responses or perpetuated stereotypes,” and “privacy issues,” and instructs employees not to enter personally identifiable information or sensitive unclassified information. How any of those warnings will be enforced was not specified.

Of course, federal agencies have been experimenting with generative AI for many months. Before the November election, for instance, GSA had initiated a contract with Google to test how AI models “can enhance productivity, collaboration, and efficiency,” according to a public inventory. The Departments of Homeland Security, Health and Human Services, and Veterans Affairs, as well as numerous other federal agencies, were testing tools from OpenAI, Google, Anthropic, and elsewhere before the inauguration. Some kind of federal chatbot was probably inevitable.

But not necessarily in this form. Biden took a more cautious approach to the technology: In a landmark executive order and subsequent federal guidance, the previous administration stressed that the government’s use of AI should be subject to thorough testing, strict guardrails, and public transparency, given the technology’s obvious risks and shortcomings. Trump, on his first day in office, repealed that order, with the White House later saying that it had imposed “onerous and unnecessary government control.” Now DOGE and the Trump administration appear intent on using the entire federal government as a sandbox, and the more than 340 million Americans they serve as potential test subjects.