We have been talking about the problem of ownership in the modern world for some time, particularly as it revolves around how digital or internet-reliant products are sold. It’s become such a prevalent problem that there’s something of a generic mantra for it: You don’t actually own the thing you bought. There’s a spectrum to this, though, which spans from the idea that digital video games are licensed rather than sold to companies either removing features sold with a product or suddenly hiding them behind a paywalled app, all the way up to companies going under and rendering products already sold to customers into useless bricks and e-waste as backend servers get unplugged.
None of these outcomes are good for the consumer, but that last category is the one that both creates the most visceral response and is the most plainly absurd. It also very much breeds distrust. You can see some of that in action when it comes to the latest example of a startup’s implosion bricking a product: Humane’s AI pin.
After launching its AI Pin in April 2024 and reportedly seeking a buyout by May 2024, Humane is shutting down. Most of the people who bought an AI Pin will not get refunds for the devices, which debuted at $700, dropped to $500, and will be bricked on February 28 at noon PT.
At that time, AI Pins, which are lapel pins with an integrated AI voice assistant, camera, speaker, and laser projector, “will no longer connect to Humane’s servers,” and “all customer data, including personal identifiable information… will be permanently deleted from Humane’s servers,” according to Humane’s FAQ page. Humane also stopped selling AI pins as of yesterday and canceled any orders that had been made but not yet fulfilled. Humane said it is discontinuing the AI Pin because it’s “moving onto new endeavors.”
Which is another way of saying that the company will be selling off all of its assets to HP and winding down completely. Hundreds of dollars have been spent on these devices by people and they’re just going to stop working. Completely. No refunds for the majority of customers, as they’re only being offered on purchases within the last 90 days. No open-sourcing of the product so that members of the public can stand up their own servers. Just… gone.
And while I will admit I struggle at times to feel a great deal of sympathy for people when they fall for a lot of the AI-hyped-up bullshit that is out there, people are pissed about this and justifiably so.
One Reddit user, for example, wrote on the Humane subreddit that they “feel like we’ve been duped.”
The announcement has also made some apparent users cynical about the intentions of the San Francisco firm, which former Apple executives launched in 2018.
“It’s truly a middle finger. Especially because there is no way around it due to the server reliance. I believe this was their plan all along. Sell and [get out],” one Reddit user said.
Similarly, another Reddit user said the lack of refunds and server access were “a blow” to early adopters, saying, “Humane won by selling. HP won a new tech. All consumers got fucked…”
The comment that sticks out to me is the one indicating that this was the plan all along. Now, I don’t believe that in its entirety. I’m certain that the folks at Humane didn’t want this exact scenario to play out. After all, last year the company was seeking a buyer willing to pay nearly $1 billion for the company. See? They didn’t get exactly what they wanted!
But creating a company to speedrun from an absurd initial valuation to a buyout by a larger tech firm isn’t exactly unheard of in Silicon Valley. Nor in many other places for that matter. And, while I doubt very much that any planning sessions at Humane involved someone saying, “And here’s where we fuck all of our customers completely by shutting down their devices, bwah ha ha!,” I have no trouble believing that this was a company created to be sold all along. And it sure does seem like the company didn’t have its own customers’ best interests in mind as it was seeking a buyer all along.
Humane also continued to push the pins despite reportedly seeking a buyer since May, and it gave AI Pin owners just 10 days to reckon with their expensive devices being bricked. In addition, the limited refund window seems like a slap in the face to people who were willing to spend extra money to be early adopters.
Here again we find that our laws simply have not kept up with the times. There needs to be some structure and rules around this sort of thing, such that the public is at least protected from buying a $700 product only to find out 91 days later that the product is gone and there is no refund coming. That such a situation is untenable is not a controversial opinion.
Maybe someone can take this up with the Consumer Financial Protection Bureau and… oh, yeah, never mind.
You might recall Buzzfeed CEO Jonah Peretti as the guy who gutted Buzzfeed’s talented news division and fired oodles of human beings back in 2023. As part of that transition, Peretti heavily embraced half cooked ‘AI’ technology in the form of generative and interactive AI chatbots he insisted would dramatically boost the site’s traffic and audience.
That didn’t do a whole lot to improve Buzzfeed’s fortunes, so now Peretti is back, with another new “pivot to video AI” that apparently involves talking a lot of shit about AI. In a new blog post, Peretti laments the way that AI has been clumsily rushed to market in a way that devalues human agency and labor, hoping you’ll apparently forget he was involved in using AI to devalue human agency and labor:
“Most anxieties about the future are really about the present. We worry about a future where AI takes away our human agency, devalues our labor, and creates social discord. But that world is already here and our meaning, purpose, and agency has already been undermined by Artificial Intelligence technologies.”
Peretti complains about something he calls SNARF, an acronym for “stakes, novelty, anger, retention, fear,” he says companies like Meta and TikTok have engaged in to grab consumer attention. Peretti’s solution to all of this? To build a new social media platform called BF Island he says will “allow users to use AI to create and share content around their interests.”
Peretti claims he’s going to be creating a “totally different kind of business, where it’s primarily a tech company and a new kind of social media company,” but it’s not entirely clear how Peretti will avoid the SNARF problem he wants you to forget he played a starring role in.
“If a lot of people click on it, it must be good” is the primary way to make money in the modern ad ecosystem, something that often directly conflicts with pesky stuff like ethics, quality, and the public interest. Peretti claims BF Island will be “built specifically to spread joy and enable playful creative expression.” Outlets like Axios can’t be bothered to mention Peretti’s role in precisely the sort of behaviors he complains about in his blog post.
Maybe Peretti can build something new and useful and interesting. But so far, AI has had a disastrous introduction to journalism and media, resulting in rampant layoffs, oodles of plagiarism, false and misleading headlines, and a whole bunch of sloppily automated news aggregation systems that are redirecting dwindling ad revenues away from real journalists and real journalism.
It hasn’t had much better of an impact on social media, given Facebook, Google, and TikTok are increasingly full of badly automated slop that’s making the internet less useful, not more.
That’s less the fault of the undercooked technology as it is the sort of fail-upward brunchlord executives in tech and media who genuinely appear to have absolutely no idea what they’re doing. The kind of folks all out of new ideas who see automation primarily as a way to dismantle labor, cut corners, save money, and create a sort of low-effort automated ouroboros that shits ad engagement cash.
Peretti very much was one of those guys, appears to still be one of those guys, yet simultaneously now wants to capitalize on the public annoyance he himself helped cultivate while very likely changing very little about what actually brought us to this point.
Here was a fun surprise last night. John Oliver just delivered what might be the most accessible and accurate mainstream takedown of content moderation myths we’ve seen yet. The latest episode of “Last Week Tonight” tackled content moderation head-on, while systematically dismantling Mark Zuckerberg’s increasingly dubious justifications for Meta’s policy changes. In this era where most mainstream coverage of content moderation is a total mess, Oliver somehow manages to both be hilarious and (surprisingly) get basically everything right about this impossibly thorny issue.
It’s worth watching, if only to see someone explain in 30 minutes what we’ve been trying to hammer home for years. (And no, I’m not just saying that because he mentions Masnick’s Impossibility Theorem — though that certainly doesn’t hurt.)
The segment hits on several key points:
First, there’s what you might call the fundamentals of content moderation (or “why the internet isn’t just porn and diet pills 101”):
Section 230 made it possible to moderate content online. Without it, websites would basically have two choices: let everything in (hello, spam!) or shut everything down. Neither is great for business, or users, or… well, anyone really.
Content moderation is an intractable issue. This isn’t just my opinion — it’s mathematics. Every platform that allows user content either moderates or dies trying. There’s no third option. (Unless you count “becoming a wasteland of porn and diet pill ads” as an option, which, fair enough, some do.)
The dirty secret is that social media companies have actually put a fair bit of effort into this problem. They’ve drawn lines, redrawn them, hired thousands of moderators, built AI systems, and… people still hate where those lines end up. Because of course they do. That’s the “impossible” part of my theorem.
Then, he debunks the false claims of political manipulation:
Oliver points out how MAGA Republicans insisting that content moderation is some sort of vast left-wing conspiracy targeting conservatives turns out to be complete nonsense.
He also does an excellent job debunking the misleading narrative around “Hunter Biden laptop” story. As we’ve written, that story has been blown totally out of proportion. The narrative says it was suppressed. It wasn’t. The narrative says the details were damning. It wasn’t that either. What it was, mainly, was a masterclass in how to turn routine content moderation decisions into political theater. And Oliver shows that clearly.
Then there’s Zuck’s latest performance piece about how the Biden administration supposedly forced him to censor content. Oliver absolutely nails why this claim is ridiculous. (Pro tip: When the government “pressures” you to do something and you just… tell them no and nothing happens in response, that’s not exactly censorship.)
And then the kicker: Oliver highlights (as we have multiple times) that even the very conservative Supreme Court has said these claims are nonsense. Though I suppose when reality conflicts with your preferred narrative, you can always just pretend the Supreme Court doesn’t exist… or that Amy Coney Barrett is too woke.
And here’s where Oliver really sticks the landing, showing where all of this is heading:
Remember all those “simple fixes” politicians keep proposing for Section 230? Oliver explains how every single one would basically hand the government (and specifically, the Musk/Trump administration) a shiny new tool to silence speech they dislike. Because nothing says “free speech” quite like giving the government more power to control online speech, right?
Finally, Oliver exposes the Zuckerberg two-step: Zuck loves to brag about how he stood up to the Biden administration’s requests, but conveniently leaves out the part where he completely rolled over for Trump’s actual threats. (You know it’s bad when Trump himself is bragging about how effectively he bullied Zuck, which Oliver points out, shows that it doesn’t take a genius to realize what really happened.)
In the end, what Oliver has given us is basically a greatest hits album of Techdirt’s content moderation coverage from the last few years, except with better production values and more jokes about Mark Zuckerberg’s new look. And the finale? A pitch-perfect “advertisement” for Facebook’s new content moderation philosophy that can be summed up in two words: Fuck It.
Blithely ignoring decades of jurisprudence, Mississippi Chancery Judge Crystal Wise Martin issued a temporary restraining order directing a small local paper, the Clarksdale Press Register to remove an op-ed that criticized the city for holding a meeting that was supposed to public without notifying the public about the upcoming meeting.
This order was crafted and issued without any input from the affected paper. Instead, the decision was made solely based on the allegations of the city and its legal representative, Melvin D. Miller II. Ken White (a.k.a. Popehat) engaged the attorney in a bit of emailed conversation, sussing out the fact that, although the attorney seems to be a decent guy otherwise, he seems to firmly believe this isn’t prior restraint because the newspaper published the op-ed before the city managed to force it to take it down.
He also seems to feel it’s “defamatory” to suggest the city purposely failed to notify the public (the city claims it was a “mistake”) — something that’s clearly not remotely true in terms of case law. And he seems to think a government entity can sue for defamation, which is similarly not even remotely true.
This is the entire op-ed that got the city so bothered it filed a patently ridiculous lawsuit and somehow managed to convince a judge to violate the First Amendment on its behalf:
Your Clarksdale Press Register will be the first to say that a sin tax that would pay police to fight crime in Clarksdale is a good idea.
So why did the City of Clarksdale fail to go to the public with details about this idea before it sent a resolution to the Mississippi Legislature seeking a two-percent tax on alcohol, marijuana and tobacco?
Mayor Chuck Espy has always touted how “open” and “transparent” he is and he is “not like previous administrations of the past 30 years.”
So why did Espy seek a Special Called Meeting of the Board of Mayor and Commissioners to finalize details of this move?
The notice was posted at city hall as required by law and said stated the city would “give appropriate notice thereof to the media.”
This newspaper was never notified. We know of no other media organization that was notified.
But back to what the city was trying to do.
Yes, there are deadlines for submitting legislation to Jackson. But this tax has been discussed in at least two meetings and has been reported in the pages of your Clarksdale Press Register.
Have commissioners or the mayor gotten kick-back from the community? Until Tuesday we had not heard of any. Maybe they just want a few nights in Jackson to lobby for this idea – at public expense.
As with all legislation, the devil is in the details and how legislation often morphs into something else that benefits somebody else.
An idea that sought to pay police higher wages for the toughest job in any community is admirable. But the way the resolution sought by the city of Clarksdale is now written gives us cause for concern.
The money – our money – can now be spent to “support and promote public safety, crime prevention and continued economic growth in the city.”
Does that mean the fire department, 911, Chamber of Commerce and their pet projects?
Does that promotion mean, giving away candy at Halloween, toy giveaways at Christmas and hosting events where politicians can hand out goody bags to voters in the name of safety?
This newspaper feels the original intent serves the purpose of all — putting police on the streets of Clarksdale.
More police will lead to more patrols, more patrols will lead to more arrests, more arrests will lead to less crime and less crime will make us all feel safer in our homes and neighborhoods.
Our Clarksdale Board of Mayor and Commissioners have stumped their toe on this one. They took a good idea, let their focus drift, and made us suspicious.
That’s being republished here because the Clarksdale paper has complied with this clearly unlawful court order and removed it from its website.
But as terrible as the city has been on this and as ignorant as its legal rep is about First Amendment law, the real villain here is Judge Martin, who could have easily determined this attempt to silence the paper was unconstitutional and told the city’s lawyer to GTFO. But she went the other way — without even giving the paper a chance to reply — and gave the city what it wanted. Here’s the most relevant part of the court order, as provided by Adam Steinbaugh:
THIS MATTER having come before this Court on the verified Petition of the City of Clarksdale requesting a temporary restraining order or a hearing on a preliminary injunction and after review of the verified Petition and Attorney Certification, the Court finds that the Petition is well taken. The injury in this case is defamation against public figures through actual malice in reckless disregard of the truth and interferes with their legitimate function to advocate for legislation they believe would help their municipality during this current legislative cycle. The Respondents are well aware of the accusations against them as they have already received a draft of the verified Petition and further notice is not needed until such hearing. The Temporary Restraining Order for Respondents to remove the article “EDITORIAL: SECRECY AND DECEPTION ERODE PUBLIC TRUST” from their online portals and make it inaccessible to the public is hereby granted.
The judge’s stance on the issue is as ridiculous as the city’s. But it looks like the city definitely wanted this particular judge to take the case. It looks like the city and its attorney did a little bit of judge-shopping before filing. There’s a Chancery court located in Clarksdale, Mississippi.
Despite that — as Steinbaugh pointed out on Bluesky — the lawsuit was filled in Hinds County, which is more than 150miles away from where the alleged “injury” took place.
Maybe city officials don’t know that law. That’s worrisome, but some lack of legal specifics is to be expected. Maybe the city’s lawyer doesn’t understand First Amendment law. That’s far more worrying, especially when a city decides it’s going to start filing defamation lawsuits that are strictly forbidden by US Supreme Court precedent. And even if the lawyer doesn’t know the law, the court — even at this lower level — definitely should. And even if Judge Martin wasn’t exactly up on all the legal precedent established… well, more than 100 years ago, she should have recognized this had First Amendment implications and, at the very least, given the paper a chance to respond before issuing a court order. Had the judge done that, there’s a good chance this order would never have been issued as the paper’s lawyers would have made it clear (1) this isn’t defamation, and (2) the city has no legal right to engage in this sort of lawsuit.
But until that happens, the op-ed has been taken down and the city is celebrating its “victory” on social media. This lawsuit should never have been filed. And no judge in this nation should have issued this order, which clearly and blatantly violates the First Amendment.
Look, there are different ways to manage people. You could, for instance, have regular performance reviews, set clear expectations, and provide constructive feedback. Or… you could send an email late on a Saturday to the entirety of the federal government workforce (even those outside the executive branch) demanding that everyone list five things they did last week, while simultaneously tweeting that anyone who doesn’t respond will be fired.
The latter is what happened this weekend when federal employees received this email:
What did you do last week?
Please reply to this email with approx. 5 bullets of what you accomplished last week and cc your manager.
Please do not send any classified information, links, or attachments.
Deadline is this Monday at 11:59pmEST.
Now, you might think this is just another story about Elon Musk’s catastrophically bad management style. (And it is!) But it’s actually much, much dumber than that.
It was sent on Saturday. And Elon Musk is taking credit for it, even though the Trump administration last week stated in court that Musk has no authority other than to advise the President, and has no official role with DOGE. Even more bizarre, Musk claimed on ExTwitter that anyone who failed to reply to the email by Monday night would have that failure to respond be taken as a resignation.
There are several problems here. Well, actually there are about fifty problems here, but let’s start with the obvious ones:
The email doesn’t mention anything about resignations. That part came in a separate tweet, because apparently that’s how the federal government works now. (If you’re a federal employee who doesn’t obsessively follow Elon Musk on ExTwitter, I guess you just… accidentally resign? Maybe?)
The federal government is, how do I put this, kind of big? Some federal employees are on maternity leave. Some are on vacation. Some are in submarines deep under the ocean where checking email would literally compromise national security. (I assume Musk would count “maintaining radio silence to avoid detection by foreign adversaries” as one of your five accomplishments for the week, but who knows?)
If this all feels familiar, it’s because we’ve seen this movie before: Musk pulled exactly the same stunt when he took over Twitter, right before destroying about 80% of that company’s value. (You would think he’d recognize how badly that has gone and think that maybe a different approach is needed, but not Elon Musk!)
Furthermore, the email went to all federal employees, including many who are not a part of the executive branch. There are multiple reports of clerks and judges in the judicial branch receiving it as well. And while we’re still waiting to see the courts sort out if Musk has authority over the executive branch (he likely does not), he absolutely does not have authority over the judicial branch.
Now, you might wonder what possible justification there could be for this bizarre demand. Well! According to Musk (who, remember, suffers from the most ridiculous level of troll-fueled confirmation bias we’ve ever seen) this is just a simple test to make sure federal employees are checking their email. Because apparently the biggest problem facing the federal government is… insufficient inbox monitoring?
There are a few problems with this theory:
Some federal employees literally can’t check email (see: aforementioned submarine crews)
Some federal employees shouldn’t check email (see: anyone handling classified information on secure systems)
Some federal employees don’t need to check email on weekends (see: basically everyone else)
But the real kicker is what Musk’s defenders are saying.
The argument goes something like this: “Actually, this is totally normal! Companies do this all the time!” Which… no? Look, I’ve worked in and around plenty of companies, and yes, you typically have regular performance reviews. You might even have weekly check-ins with your manager. But there’s a slight difference between “scheduled performance review with your direct supervisor” and “surprise email from someone who may or may not have authority over you demanding immediate justification for your existence.” (The difference is that one is management and the other is performative chaos.)
That’s just being an asshole with too much power.
Also, because these are federal government emails, they’re subject to the Freedom of Information Act, which means reporters are already lining up to request copies of all the responses. I suspect we’ll soon have a fascinating database of federal employees explaining their jobs to… well, to no one in particular, since Musk doesn’t actually have any actual authority here.
Not surprisingly to most people, but apparently surprising to Musk, it turns out that various federal agencies have opinions about their employees sending detailed work descriptions to random email addresses. And those opinions are mostly variations on “please don’t do that.”
The FBI, for instance, whose new director Kash Patel (in theory a Musk ally, mind you) seems particularly annoyed:
Then there’s Tulsi Gabbard, the new Director of National Intelligence (and, again, typically a Musk ally), who had to explain something that really shouldn’t need explaining: “Given the inherently sensitive and classified nature of our work, I.C. employees should not respond to the OPM email.” (Translation: “Please don’t send classified intelligence work details to a random email address, even if Elon Musk asks nicely.”)
The Defense Department, meanwhile, sent out what might be the most diplomatically worded “absolutely not” in recent memory. From their memo:
“DoD personnel may have received an email from OPM requesting information. The Department of Defense is responsible for reviewing the performance of its personnel and it will conduct any review in accordance with its own procedures,” Selnick wrote. “When and if required, the Department will coordinate responses to the email you have received from OPM. For now, please pause any response to the OPM email titled, ‘What did you do last week.’”
The Administrative Office of the Courts, which is run by John Roberts, sent out a mealy-mouthed email to the judicial branch recommending not responding: “this email did not originate from the judiciary or the administrative office and we suggest that no action be taken.” Roberts could have taken a stand and noted that the executive branch has no authority whatsoever here, but I guess he’ll have an opportunity to do that in court before long.
The State Department and Homeland Security both also told employees not to respond. Though CISA, which is a part of Homeland Security, first told employees to obey the email. That kind of confusion is happening elsewhere as well:
Other departments gave conflicting guidance. The Department of Health and Human Services told its employees on Sunday morning to follow the directive. An hour later, an email from the Trump-appointed acting director of the National Institutes of Health, a subordinate agency, told employees to hold off on responding. Hours later, the health department told all employees to “pause” responses to the ultimatum.
One message on Sunday morning from the Department of Health and Human Services, led by Robert F. Kennedy Jr., instructed its roughly 80,000 employees to comply. That was shortly after the acting general counsel, Sean Keveney, had instructed some not to. And by Sunday evening, agency leadership issued new instructions that employees should “pause activities” related to the request until noon on Monday.
“I’ll be candid with you. Having put in over 70 hours of work last week advancing Administration’s priorities,I was personally insulted to receive the below email,” Keveney said in an email viewed by The Associated Press that acknowledged a broad sense of “uncertainty and stress” within the agency.
Keveney laid out security concerns and pointed out some of the work done by the agency’s employees may be protected by attorney-client privilege: “I have received no assurances that there are appropriate protections in place to safeguard responses to this email.”
Look, even if you were somehow convinced this was a good idea (it’s not) and that demanding work summaries via surprise weekend email is totally normal corporate behavior (it really, really isn’t), you’d still have to marvel at the sheer incompetence of the implementation. All this is doing is generating a shit ton of confusion across the entirety of the federal government.
That doesn’t seem very useful for “efficiency.”
And then there’s Ed Martin, the US Attorney for DC (who, you might remember, we just last week discussed as spectacularly incompetent), who sent what might be the most confusing “clarification” email in federal government history:
“Let me clarify: We will comply with this OPM request whether by replying or deciding not to reply.”
Well! That certainly clears things up. (For those keeping score at home, Martin is saying they will comply by either… doing the thing or not doing the thing. Which is technically true and also technically useless.)
But wait, there’s more! Because Elon (who, remember, is supposedly just an advisor with no actual authority) didn’t take kindly to the Pentagon’s “please ignore this” memo. His response? To threaten to fire the person who wrote the Pentagon’s memo. Yes, the person with no authority is threatening to fire people at the Pentagon for not recognizing his non-existent authority. It’s like a fractal of nonsense.
Meanwhile, Musk has been gleefully mocking anyone pushing back on this demand, insisting that people are only upset because they can’t come up with five things they did last week. Which is… not the point. At all.
Let’s be clear about this (in “five bullets”):
Everyone can list five things they did last week
The issue isn’t the difficulty of the task
The issue is being asked to justify your existence via a pointless busywork exercise to someone with no authority over you
…via a weekend email
…that threatens termination in a separate tweet
But Musk wasn’t done yet. Because his solution to this manufactured crisis is… wait for it… to use his own proprietary AI chatbot to generate fake responses. Yes, you read that right. Musk sent Trump a screenshot of someone (possibly himself) asking Grok (his own AI) to make up fake accomplishments for such an email reply, which Trump then posted to Truth Social, which Musk then reposted to ExTwitter as proof of how “easy” this all is.
So to summarize: The person demanding accountability from federal workers is actively encouraging them to use AI bullshit generators to create fake responses. And not just any AI — his AI specifically. (Nothing says “government efficiency” quite like using a private company’s AI to generate fake work reports for that same private company’s CEO who has no actual government authority but pretends he does.)
It also suggests a disturbing comfort with using AI to generate artificial accountability rather than pursuing any kind of meaningful government oversight (in case you were one of the three rubes left in the country who still believes that’s what Musk is doing). The fact that neither Musk nor Trump seem concerned about the security implications of federal employees feeding their work details into private commercial AI systems is particularly alarming.
There are a whole host of problems with all of this, but mainly, it’s just fucking stupid.
Kelley said in the letter that the union has “received numerous reports from dedicated civil servants, including those who care for our veterans and safeguard our nation, expressing frustration over the email’s tone and intent. Rather than fostering professionalism and respect for their work, this hastily written email left many feeling undervalued and intimidated.”
And even Republicans are having trouble defending this one.
Senator Lisa Murkowski, Republican of Alaska, also criticized Mr. Musk’s order.
“Our public workforce deserves to be treated with dignity and respect for the unheralded jobs they perform,” she wrote in a statement on social media. “The absurd weekend email to justify their existence wasn’t it.”
The whole thing is an exercise in dickishness for the sake of dickishness. But beyond the obvious management failures, this episode raises serious concerns about data security and privacy. The combination of FOIA-able responses, encouraged use of commercial AI systems, and the broad scope of affected agencies creates a perfect storm of potential security risks. Federal employees’ work details could be exposed in ways that compromise ongoing operations, especially in sensitive areas like national security and law enforcement. It’s yet another example of how tech-bro solutions to imagined problems often create very real security vulnerabilities.
Of course, Musk fans will cheer it on, insisting that the federal workforce deserves to be treated like shit, even as this will impact many people who actually supported Trump and Musk. The entire attitude is “if you’re not part of the inner circle, you’re worthless.”
It’s obnoxious. And it’s designed to demoralize workers on purpose. The assumption that all federal employees are a waste is such a stupid, ignorant position. But it’s clearly how Musk is treating everyone who works for the government.
StackSkills is the premier online learning platform for mastering today’s most in-demand skills. Now, with this exclusive limited-time offer, you’ll gain access to 1000+ StackSkills courses for life! Whether you’re looking to earn a promotion, make a career change, or pick up a side hustle to make some extra cash, StackSkills delivers engaging online courses featuring the skills that matter most today, both personally and professionally. It’s on sale for $19.97 for a limited time.
Note: The Techdirt Deals Store is powered and curated by StackCommerce. A portion of all sales from Techdirt Deals helps support Techdirt. The products featured do not reflect endorsements by our editorial team.
This post was written on Saturday before news broke that Elon Musk had commanded every single federal employee—including those in the judiciary!—to send a “five things I did last week” email to hr@opm.gov. But even that episode, where Musk and DOGE once again flexed power they don’t lawfully have, and in contact with computer systems and data that they don’t lawfully have access to, just serves as yet one more example of the issues this post was written to discuss.
This post talks about a few things: (1) a small but important win that New York got on Friday, when a court again recognized that what DOGE has been allowed to do at the Treasury department is nuts (and basically illegal), and (2) some issues that are emerging when it comes to demonstrating the standing needed to sue for all the destruction being wreaked across the Executive Branch, why more courts need to recognize that they are not issues that should be derailing these cases, and why all these cases are still, at their core, about Musk and DOGE’s unlawful intrusion into the nation’s most sensitive computer systems.
First the New York case: on Friday the court in New York v. Trump turned the TRO that had been limiting what DOGE could do in the Treasury department into a preliminary injunction. In doing so the court found that the plaintiff states, led by New York, had shown a likelihood of success on their claim that the Treasury department was “arbitrary and capricious” in letting DOGE tear through its systems (and thus violated the APA):
Based upon the factual record developed to date, the Court finds that Plaintiffs will more likely than not succeed in establishing that the agency’s processes for permitting the Treasury DOGE Team access to critical BFS payment systems, with full knowledge of the serious risks that access entailed, was arbitrary and capricious. While it appears that the career staff at BFS did their best to develop what mitigation strategies they could, the inexplicable urgency and time constraints under which they operated all but ensured that the launch of the Treasury DOGE Team was chaotic and haphazard. […] The record is silent as to what vetting or security clearance process [Krause and Elez] went through prior to their appointment. […] The Treasury DOGE Team started its work almost immediately, even though it did not yet have either the HR specialist or the attorney that the E.O. mandated should be members of the team. This left career staff with almost no time to develop their mitigation measures. Within days of [Elez’s] appointment, and apparently after receiving minimal, if any, training regarding the handling of sensitive government information (beyond being instructed to maintain the information on his BFS laptop), Elez was given full access to system source codes. […] Even now, weeks after his departure, the Treasury Department is still reviewing his logs to determine what precisely he accessed and what he did with his access. The Treasury Department also could not confirm whether or not Elez emailed PII or other confidential information to officials outside the Treasury Department.
And then there is the question of under what authority anyone from DOGE had any access at all:
It is also unclear from this record whether the agency established clear reporting lines for the Treasury DOGE Team. Although they are nominally agency employees who sit within the Treasury chain of command, it is notable that they also take instructions from officials at USDS/DOGE. How this works in practice, and the uncertainty this creates as to their status as Treasury employees, calls into question their authority to access Treasury record systems.
So with this new injunction the Treasury department has now basically been ordered to run its department as the law requires, including in how the law requires restricting computer access to certain personnel and only after they have been onboarded properly and with the appropriate vetting. And from the outset that means no one connected with DOGE qualifies:
[The] United States Department of the Treasury and the Secretary of the Treasury are restrained from granting access to any Treasury Department payment record, payment systems, or any other data systems maintained by the Treasury Department containing personally identifiable information and/or confidential financial information of payees to any employee, officer or contractor employed or affiliated with the United States DOGE Service, DOGE, or the DOGE Team established at the Treasury Department, pending further Order of this Court[.]
If the Treasury department wants to give them access, then it’s going to have to do what the law requires before they can, and the court is going to make sure it does:
[B]y Monday, March 24, 2025, the United States Department of the Treasury shall submit a report to this Court: (i) certifying that the Treasury DOGE Team members have been provided with all training that is typically required of individuals granted access to BFS payment systems, including training regarding the federal laws, regulations, and policies governing the handling of personally identifiable information, tax return information, and sensitive financial data, and maintaining the integrity and security of Treasury data and technology, and attesting that any future Treasury DOGE Team member will be provided with this same training prior to being granted access to BFS systems; (ii) certifying the vetting and security clearances processes that members of the Treasury DOGE Team have undergone, and how that vetting process compares with the processes undergone by career employees who have previously been granted access to the BFS payment systems; (iii) describing the mitigation procedures that have been developed to minimize any threats resulting from increased access by members of the Treasury DOGE Team to BFS payment systems; (iv) setting forth the legal authority pursuant to which each DOGE Team member was employed by or detailed to the Treasury Department; and (v) explaining the reporting chains that govern the relationship between the DOGE Team members, USDS/DOGE, and Treasury leadership (with reference, if applicable, to any Memorandum of Understanding setting forth that relationship).
This injunction is a pretty good, albeit narrow, result. It’s good because it gets the job done: it gets DOGE out of the Treasury department, at least in the dangerous, unaccountable way it had been. And it’s good because it is yet more judicial recognition of how dangerous and unaccountable DOGE has been, and in a way that law was unlikely to allow.
The injunction is narrow, however, because the court also rejected most of the claims the plaintiff states had brought, and also many of their claims of standing. And while this rejection doesn’t matter here—in some cases the rejection was probably reasonable, and in any case one claim did stick sufficiently, which was all that was needed—some of the analytical issues that this court struggled with are also tripping up other courts, especially when it comes to them finding the standing needed to give plaintiffs the injunctive relief they need.
All of this litigation we are seeing is something of a jigsaw puzzle, one whose picture is slowly coming together, with lots of different pieces, including a variety of plaintiffs, a variety of claims, and a variety of defendants, and even types of defendants. For instance, we’ve been tracking when Musk and DOGE themselves started to be directly named as defendants, partly because we want them to ultimately be held directly liable for the damage they are causing, but also partly because, for any lawsuit seeking to remediate (or enjoin) a harm, the argument for who was supposed to stop the harm, and why, is different depending on who the defendant is. For instance, it is different to sue an agency and its head for what the agency has done wrong (ex: allowing DOGE to mess with its systems, because it was beyond their own power to allow it) than it is to sue Musk or DOGE for what they are doing wrong (ex: messing with those systems).
In some cases, like AFGE v. OPM, we are seeing a hybrid, and it’s starting to seem like the hybrid approach may be the way to go forward in most cases because it covers both bases and paints a more complete picture of what is going wrong and why injunctive relief is necessary to stop it, and proper to award. Not every court has been convinced, like in AFGE v. Trump, one of the earliest cases to be filed, and one challenging DOGE’s efforts to destroy USAID. Although a TRO was initially granted, and it provided some interim relief, it was only temporary. Last week the court dissolved it and declined to grant a preliminary injunction to keep the agency from trying to fire their own workforce. In declining to grant the sought injunction a significant part of the reasoning was that terminations are normally properly adjudicated via specialized agencies that Congress has established, and that job loss itself wasn’t “irreparable” enough.
But the upshot of this decision is that the court has basically thrust the unions and their employee members into a Kafkaesque nightmare where employees need to be fired first, at which point they will then need to take their claims for wrongful termination (presumably one-at-a-time, instead of collectively as a union), to a different agency tasked with arbitrating federal employment disputes (and which Musk and DOGE have already started to dismantle) to try to get their jobs back. Only then, after this avenue has been exhausted, do USAID employees have any chance to get back to Article III courts to address any of the unconstitutional illegality underpinning the firings and destruction of their agencies in the first place.
Which can’t possibly be the right result, because it would effectively leave them without a remedy for their wrongful termination. The problem is, the statutory scheme that the courts are pointing to, which Congress created to shunt employment disputes to, only makes sense in the context of normal agency operations, which cannot lawfully include dismembering themselves without Congressional authorization. And it certainly cannot include dismembering themselves at the direction of an entirely unlawful power like DOGE or Musk, because of course there were also laws designed to prevent this illegal situation from ever arising. (See for instance recognition by the court in AFL-CIO v. Department of Labor that DOGE may be an improperly-formed agency given the Economy Act of 1932).
What is happeing now if far beyond any sort of HR dispute; it’s the lawless dismantling of government agencies Congress established by law, of which there are myriad consequences, only some of which are experienced in terms of employment. But the goal of all these lawsuits is not just to save the jobs, as in this case, or data privacy, as in others, but to restrain the lawlessness that is causing any of these things (and so much more) to be lost. And addressing that lawlessness is absolutely the purview of the courts—in fact, because they are not Article III courts it may even be inappropriate for these other employment-related agencies to try to address it themselves. These employment-related agencies may ordinarily be able to help employees keep their jobs when an individual firing may be beyond what the law would allow an agency to do, but they can’t fix the real problem that is causing the firings here, at the behest, if not also directly at the hands, of people with no lawful authority to compel them. And it would lead to a bizarre result if Musk and DOGE, as people acting so far beyond the bounds of lawful authority, could then be protected by an actual law now preventing them from being held accountable for it.
In the USAID example, the real problem is not that this agency has suddenly, and independently, decided to destroy itself but that its destruction is clearly being directed by Musk and DOGE, who have no lawful power to do so. They have openly, and repeatedly, bragged about putting the agency into the “woodchipper.” And, as the court in New Mexico v. Musk observed, Musk and his DOGE minions seem to be the supervisory authority governing every contract cancellation, funding freeze, and firing at any agency they have been involved with.
In light of this increasing judicial recognition, Musk and DOGE should from now on probably be named in just about every lawsuit brought to challenge what is happening in the Executive Branch, even if the agency and its real officials are also named, because it appears to be Musk and DOGE’s ultra vires behavior that is at the root of all the harm accruing. It also seems important to name them given all the vagueness and inconsistencies in the government’s declarations about their behavior and what authority it now claims to have behind it, which courts are starting to call out as we also saw in New Mexico. With these declarations it seems like the Trump administration has apparently begun to retroactively try to dot some of the i’s they should have had dotted before Musk and DOGE started acting so radically, like properly hire and vet staff that gets to access Treasury’s computer systems. But it does seem like these efforts are too little too late: even in the case of the Treasury department, discussed above, even if the agency now properly hires all the DOGErs, unaccountable DOGE personnel had still been mucking about in those computer systems for way too long without those formalities being satisfied, and in doing so creating exactly the sort of problems that those formalities were supposed to forestall. Just as they have in every other agency they’ve invaded.
And ultimately it seems like pretty much all the resulting harm being sued over—contract cancellations, funding freezes, or firings—originates from Musk and DOGE’s incursions into the agencies’ computer systems. Which is important to explain to courts, for several reasons.
One is with respect to an issue that has started to come up in some of these lawsuits, addressing how a case called TransUnion applies. TransUnion is a case about “standing.” In general people can only sue when they have standing, or, in other words, an actual (or very likely) injury that is redressable if the court were to give them the relief they want from this particular defendant. So one thing the Trump administration has been trying to do to dismiss all these lawsuits is argue that the plaintiffs bringing them don’t have the standing needed to demand the injunctions they are demanding. And one argument they’ve used in some of the cases addressing the privacy harms resulting from DOGE running rampant through these systems, is that the plaintiffs don’t have the standing that TransUnion says they need to have to complain about the privacy harm DOGE’s actions may have caused.
In TransUnion the Supreme Court wouldn’t let concerns that impermissible data access might lead to harm give the plaintiffs standing to sue for the data breach itself because the possible harm was just too hypothetical. So what the Trump administration is arguing, and come courts, like in EPIC v. OPM, are accepting, is that any worry about data DOGE may have exfiltrated potentially falling into the wrong hands is worry about a harm too hypothetical to entitle anyone to sue over it. In other words, “Yeah, the bad guys might get your data, but they might not, so no standing for you.”
But what this argument misses is that DOGE itself are the bad guys! The bad guys already got the data! Their unauthorized access to it was the exfiltration! The fact that yet more bad guys may also get the data is beyond the point. DOGE’s penetration into these secure systems, gaining access to data that was supposed to be protected against unauthorized access, for all the reasons that it needed to be kept secure from unauthorized access, is why all the people who are now suffering a consequence from that unauthorized access—including their sudden loss of agency employment, without the authorization of Congress—should now have standing to sue to stop that harm.
Because it was only because of that access that such consequences are accruing. Through their unauthorized access to all these agency systems Musk and DOGE got the visibility they needed to be able to direct all the contract cancellations, funding freezes, and firings that they have already directed or yet plan to. We know there is this connection between their destructive demands and the access to these systems they’ve had because they have essentially publicly claimed as much, and because they would not have needed to demand access to these systems in the first place if they could have done their damage to the agencies without it. Even to the extent that agency officials may now seek to launder DOGE’s unconstitutionally destructive demands (like to fire the majority of agency personnel) by implementing them under their own auspices, such ratification of these inherently unconstitutional plans is irretrievably tainted by DOGE’s interference, which was enabled by their illicit intrusion into these agencies’ protected computer systems. That unlawful intrusion was the predicate act from which all the subsequent harms have flowed, and any lawsuits challenging any of this resulting harm probably needs to make that reality prominently clear—and before an Article III court that should be ready to stop it.
The 2021 American Rescue Plan Act (ARPA) continues to quietly help fund a number of extremely popular community-owned, open access fiber deployments that are challenging entrenched U.S. monopoly power, and driving super cheap, community-owned and operated fiber networks into long neglected towns.
New York State, for example, just leveraged ARPA funds to give a $26 million grant to Oswego County. Oswego County is going to use that money to build an open access fiber network. That means multiple ISPs can come in and compete over shared infrastructure owned by the county. Our Copia report showcased how this model can help disrupt monopoly power and lower broadband costs for users.
The anchor tenant on Oswego County’s new network, Empire Access, will provide locals with 500 Megabit per second (Mbps) service for $50 a month; symmetrical 1 gigabit per second (Gbps) service for $65 a month; and symmetrical 2 Gbps service for $100 a month.
That’s not great news for regional New York State monopolies Charter and Verizon, who’ve grown fat and comfortable charging much higher prices for much slower access. The lack of real competition between the two giants for decades has resulted in high prices, slow speeds, spotty coverage, inconsistent upgrades, repair delays, and substandard customer service.
Meanwhile in Minnesota, Carver County officials say they’ve also been leveraging ARPA funds to deploy affordable gigabit fiber to every county resident. Their model is slightly different: The city has used grant money to build dark fiber, which they then lease to a company called MetroNet as part of a public-private partnership. MetroNet is offering locals gigabit fiber for prices way less than regional monopolies:
“Metronet currently offers four tiers of service with varying promotions, which currently include symmetrical 150 megabit per second (Mbps) fiber for $35 a month; symmetrical 500 Mbps for $45 a month; symmetrical 1 gigabit per second (Gbps) for $50 a month; symmetrical 2 Gbps for $70 a month; and symmetrical 5 Gbps for $110 a month.”
Again, this kind of stuff doesn’t get much attention from a press that declares infrastructure too boring to cover. But this kind of stuff is quietly transformative all the same. It’s also not clear to me why Senate Democrats aren’t competently messaging the impact ARPA funds are having on affordable broadband. Or local community centers, local road improvements, or affordable housing.
But some states (most notably Vermont, Maine, California, and New York) are trying a different tack: they’re investing heavily in community-owned open access infrastructure, and treating broadband more like an essential utility (where maximizing shareholder profits isn’t the top priority). They’re leveraging an historic infusion of federal funds to put local communities in charge of their own connectivity fate.
Entrenched telecom monopolies, which have worked tirelessly over decades to dismantle broadband competition and state and federal oversight, have worked tirelessly to demonize and undermine community broadband access. But in a decade it should be interesting to see what the data says about the differing approaches.
Keep in mind that states are also poised to receive more than $42.5 billion in additional broadband grants courtesy of the 2021 infrastructure bill. That program has significantly more restrictions than ARPA, and there’s every indication that the Trump administration will do its best to redirect as much of that money as possible away from community owned endeavors and toward companies that kiss Trump’s ass.
