Watch: Elon Musk is trying to bail out Republicans while flooding X with lies. Election Day is less than three weeks away, and panic is setting in for the GOP. “We see Republican groups pulling money away from House incumbents,” says host David Corn on the latest episode of Stress Test. “They’re cutting them off. They’re saying, ‘So long, Charlie! We don’t think you’re going to survive this blue tsunami.’” The one lifeline Republicans have right now is big money—and nothing gets bigger than Elon Musk, who has poured $366 million into this election cycle in an attempt to rescue at-riskRepublicans, making him the biggest known political donor in midterms history. But David warns, “It’s not just money with Elon Musk. It’s also disinformation.” On this episode, David and fellow host Garrison Hayes check the state of the race as we hurtle toward November. They’re joined by senior reporter Anna Merlan, who has been tracking Musk-amplified disinformation attacks on Texas Democratic Senate candidate James Talarico, who’s been the target of a MAGA operation that falsely claimed he left the campaign trail for a secret stay in an Ohio mental hospital—as well as a homophobic smear from Vice President JD Vance. “This new stuff about him being gay and mentally ill—that’s an escalation,” Anna says. And she’s watching for more: “What tends to happen on X.com these days, especially after Elon Musk took it over, is that there is a lot of disinformation in October leading up to elections—purported leaks of consequential information.” Meanwhile, Garrison and David discuss the epic fall of Kimberly Guilfoyle, ousted as U.S. ambassador to Greece after a report that she asked a donor to pay off a $100,000 Amex bill in return for providing him access to administration officials, and the shocking DHS raid on Minnesota Rep. Ilhan Omar’s Washington, DC., home. They explore the Trump administration’s pattern of targeting prominent Black women and what David dubs Trump’s “personal vendetta” against Omar. You’d think a president staring down lame-duck status would have bigger worries than settling scores. After all, reality on the campaign trail is biting hard: “Trump is running around the country trying to shore up Republicans in states where he should not need to be campaigning,” David says. All that, and your audience questions answered, on this Friday edition of “Stress Test,” our special midterms video podcast. Subscribe: youtube.com/motherjones. This transcript was generated with automated tools and may contain errors. David Corn: Elon Musk, a disinformation agent. Garrison Hayes: Trump sics the DOJ on a Democrat. Kimberly Guilfoyle— DC: What a story. GH: Yeah, and is it time for the GOP to panic? DC: This is Stress Test from Mother Jones. I’m David Corn. GH: And I’m Garrison Hayes. Welcome back to the fifth episode of Stress Test. We are three weeks away—can you believe it?—from the midterms and approaching the midpoint of this kind of pop-up video podcast that we’ve been on. So we thought it’d be a good time to come back together and to try to think about, like, what’s going on, do a little bit of a status check on how things are going. David, what’s on your mind? DC: Well, we see that the polls are not changing. They’re all moving in the same direction. They’re all benefiting Democrats, and it seems kind of baked into the cake. Trump is running around the country trying to shore up Republicans in states where he should not need to be campaigning. He went all the way to Alaska, and he took half the cabinet with him because the Democrats are looking good in the Senate race there. And we see Republican groups pulling money away from House incumbents, Republican House incumbents. They’re cutting them off. They’re saying, so long, Charlie, we don’t think you’re going to survive this blue tsunami. So people like Thomas Kean Jr., who’s a representative in New Jersey, and Mariannette Miller-Meeks, who’s in Iowa, they’re cutting them off, leaving them to their own devices. The one lifeline the Republicans have right now is big money, and the big news in big money this week, Garrison, was Elon Musk, right? He’s putting in 300—so far, $366 million into this campaign, making one of the largest donations ever in the history of American politics. I think he’s gonna top half a billion dollars. It’s going to Senate races. It’s going to get-out-the-vote campaigns and for senators and Republicans in the swing states, the key states. And it’s not just money with Elon Musk. It’s also disinformation. GH: Indeed it is. And so big money, Elon Musk, disinformation—to talk about all of this, we’re going to have a good conversation with our senior reporter here at Mother Jones, Anna Merlan. Anna, welcome to the show. Anna Merlan: Hi. Thanks for having me. DC: Yeah. I mean, everybody this week is going on and on about the money he’s pouring in—$158 million into Senate races—but you had a different take. Not only is he pouring money into this campaign, he’s a major force of disinformation. Explain. AM: Well, so this week we saw Elon Musk engaging pretty heavily with a story about Texas Senate candidate James Talarico. There was some scuttlebutt on the right because Talarico had not been seen on the campaign trail for a couple of weeks. That’s because he had the flu. He was really clear that he had the flu. He was appearing remotely via Zoom at several campaign events because he had the flu. Instead, a couple of far-right, self-styled journalists started to make a different claim: that he had actually been secretly committed to a mental hospital in Ohio for an extraordinarily short period of time before reappearing on the campaign trail. So this story didn’t work for a bunch of reasons that we can get into, but one person who was very excited about it was Elon Musk. He was engaging directly— DC: So what did he do? AM: He was engaging directly with Nick Sortor and Rhein Amacher, who are the two self-proclaimed MAGA-leaning journalists who were promoting this story. When they released a purported phone call between one of them and the mental health treatment center they claimed Talarico was at, he responded with two exclamation points. He also said at least twice in tweets that the Talarico campaign was hiding something that voters deserved to know about. So it’s pretty obvious that he had probably become aware of this story prior to Sortor and Amacher breaking it publicly. So he was a primary driver of people seeing this story. I mean, Sortor has a pretty good audience, but nothing like what Elon Musk has on the platform that he owns. GH: I’d love to get into why this story doesn’t work, because often, Anna, when these conspiracy theories come up, they’re based in some kind of thing that gives people a clue that there might be something going on. And so there are kind of this hodgepodge of facts that are unrelated, and they pull them together to make a story. But why are you saying that this story doesn’t make any sense? AM: There is a really basic timeline issue, which is that supposedly Talarico, according to Sortor and Amacher’s self-proclaimed reporting, had been released from this mental hospital in Ohio on Saturday night. That was a really big issue because he had been seen campaigning in person in Texas on Friday and Saturday. So they were crowing about a call that they had with this mental health treatment center in Ohio where, after asking someone at the facility several times if somebody by the name of James Talarico was there and getting several denials, eventually the person on the other end of the phone says, oh, he was discharged Saturday night. This person doesn’t sound clear, and this is after what turned out to be about six minutes of them asking this question in various ways. But regardless, that’s not how time works. If you see him in Texas on Friday, he cannot be in mental health treatment on Saturday. The other thing that happened is that the mental health treatment center, the Lindner Center, tweeted, James Talarico has never been treated here. Like, they were more than willing to just say, this is not real, this is not true. And this happened immediately. But that was apparently not enough for some of the people involved in the story to declare it not to be true. DC: Well, you know, it’s interesting because this is a mental health attack that they’re trying to put on to Talarico, and the other things that they’ve gone after him for, claiming he’s a vegan when he’s not—not that there’s anything wrong with that. And also, we saw what came from our own vice president, JD Vance, this week. Let’s look at that video and how he whipped up an attack against Talarico. [Clip plays.] DC: So he’s not only mentally ill and a vegan, apparently he’s gay. Wow. It seems to be the best they have at the moment. AM: Well, so there’s been a lot of this type of thing about Talarico, and some of it, again, Elon Musk has engaged in. There have been a bunch of quotes recirculated, for instance, about him denouncing white supremacy during 2020 and expressing support for the Black Lives Matter movement, him talking about, you know, trans inclusion in sports, him talking about the ills of capitalism—things that are taken to be extremely frightening by the right. And there’s been a lot of recirculating of those quotes. But this new stuff about him being gay and mentally ill, that is—that’s an escalation, I would say. DC: And we talked earlier about the money he’s putting in—10 million—tens of millions of dollars into get-out-the-vote efforts in Texas, and a lot of those are focusing on these sort of issues, claiming that Talarico has a radical trans agenda. AM: Correct. I mean, part of what’s interesting here is that after Elon Musk spent millions of dollars helping to get President Trump back into office, he said he was going to pull back on political spending. That was probably the result of a couple of things. One was that he put a lot of money into the Wisconsin Senate race, which didn’t turn out the way that he wanted it to be, including at one point he offered to essentially pay voters to vote, which you’re not allowed to do. He had to walk that back very quickly because it is illegal. And the other thing that happened was that he put a lot of support behind Germany’s far-right AfD party, even appearing to speak at an AfD rally, and during that election, again, AfD didn’t do particularly well. So in 2025, he said, you know, I’m going to pull back on political spending, I’ve done enough. But now here he is again, you know, a year later, getting more involved in the midterms, I think, than any other single donor ever has in history. GH: Yeah, and I guess the question that everyone’s wondering is, will this work? I mean, obviously he’s had some failures in the past, but do you think, here we are in this position again where Elon Musk is dumping a bunch of money into an election, do you anticipate that it’ll work this time? AM: I don’t actually think there’s a direct 1-to-1 predictor between amount of money and success, which I do think is what Elon Musk thinks. I think he assumes that a large enough amount of money will result in the outcome you want, but he should know that that’s not the case. That’s not what happened in Wisconsin. That’s not what happened, you know, with him appearing to speak on behalf of AfD. Like, it just doesn’t seem to work that way. But I guess we’ll see. And, you know, I guess he’s welcome under the laws of this country to spend as much money as he wants trying. DC: Well, Anna, we ask this question of all our guests: What’s stressing you out, or what are you watching for in the next few weeks before the election hits Election Day? AM: I feel like if I let anything stress me out at this point, I would be dead, so nothing. Just, I’m good. I will say I am paying a lot of attention to purported October surprises. You guys both know that before consequential elections, there’s often this thing termed the October surprise, a last-minute, consequential piece of information. What tends to happen on X.com these days, especially after Elon Musk took it over, again, is that there is a lot of disinformation in October leading up to elections—purported leaks of consequential information. You might recall, like, the Macron leaks that were attempted to float before a French presidential election a few years ago on Twitter that didn’t quite work. So I am paying a lot of attention to narratives that take hold on X in the next few weeks, especially things that Elon Musk is promoting. For instance, right now he’s very engaged with allegations about Ilhan Omar—either, you know, hiding the source of her wealth, you know, changing her last name. There is a lot of attempts right now to just kind of attach general suspicion to her in the lead-up to the election. So I would say that’s the primary thing I’m looking at today, but who knows what it’ll be next week. GH: Anna Merlan, thank you so much for joining us. We’ll definitely be paying very close attention to your reporting, and we appreciate you being here. AM: Thanks for having me. Thank you guys for your work. DC: Well, she mentioned Elon Musk going after Ilhan Omar, and this week, in what was a shocking move, the investigators from DHS, the Department of Homeland Security, raided the Washington home of Democratic Rep. Ilhan Omar, apparently looking for material relating to immigration forms filled out years ago, maybe even tens of years ago. This is unprecedented, Garrison, because, as you know, there’s a Department of Justice guideline that acts like these—investigations, raids, subpoenas—should not come so close to an election unless they are unavoidable. There’s a danger present. Someone may run away to, you know, flee the cops, go to another country. If there’s no reason to move this quickly before an election, so close to an election, you basically lay low. But the Trump administration is doing the opposite here. GH: They’ve, of course, you know, been targeting Congresswoman Omar for a while now. It’s a part of a larger anti-immigrant thing that they’ve been on. Obviously, we know Ilhan Omar represents a congressional district in Minnesota, and so all of this is connected to this larger project that the Trump administration has been pushing for quite some time now. But we also know that Ilhan Omar is a member of the Congressional Black Caucus. She is a Black woman. And there have been a number of these kinds of targeted attacks from this administration on Black women. We have the Fed Gov. Lisa Cook. We, of course, have Letitia James in New York City. We have LaMonica McIver, who is facing 17 years for her interactions with ICE just last year. And so this is something that’s kind of been going on for a while, David, and it does feel in some ways unprecedented, but it also feels like something that they’ve been pushing. It feels very related to the objectives of this administration. DC: Yeah. And one small point that I see here, because I just came out with this book, How Russia Won, about the Russian intervention in the past elections to help Donald Trump, and in 2016, the FBI was conducting an investigation of interactions between the Trump campaign and Russia throughout the 2016 general election period. And they were assiduous about not issuing subpoenas, not even bringing in people to interview, because they thought that might tip off the investigation to the public, and they did not want to affect the election in that way. The deep state, as Trump would say, was actually protecting the Trump campaign, and that’s the way it’s supposed to work. Of course, at the end of that campaign, James Comey revived the investigation of Hillary Clinton, and that had an impact on the campaign, and that’s the reason for this rule that they are violating. But Trump, you know, has made, you know, this crusade against Somali Americans a tremendous imperative for him—targeting her, but just the whole idea that Somali Americans are committing fraud, that they’re evil, terrible people. He said that, and they should all be shipped back to Africa. So this is racism. It’s a personal vendetta, and what it does, I think, also is just send a message that we’re in favor of cruelty, and if you get on the wrong side of us, we will come after you. GH: Yeah, it is, I think, a lot of those things. As I said before, there is this kind of personal fixation that the president has with Minnesota. To your point, the story around Minnesota, you know, quote-unquote, fraud—some of it legitimate, some of it illegitimate. We know that that’s all kind of a part of this story as well, that we’ve seen play out for at least the last nine, 10 months. But really, it kind of goes back. We know, of course, Ilhan Omar was a part of that kind of Squad group of folks elected to Congress during Trump’s first term. She has been an unapologetic opponent of the president, quite antagonistic and unapologetic in her kind of prodding of the president. So all of that feels deeply, deeply connected. David, I’m curious, what is Omar’s reaction to all of this, to her home being raided? How is she responding to this? DC: Well, she’s leaning into it. I mean, I’ve gotten many fundraising letters or emails from her. GH: Never miss an opportunity to raise a few bucks off of this stuff, right? DC: Yeah, don’t put a crisis to waste, ever. And so she’s, you know, defying Trump, standing strong, and using it to try to bolster her support. And indeed, this is, you know, elevating her. I don’t think there’s a question about her winning her election. But it is, again, you know, this is happening through DHS, and there are people who are worried about DHS agents or ICE agents showing up at the polls to try to dissuade people from being there, creating crisis, creating chaos, which is what Trump is looking to do. And, you know, they keep trying to bring this back to fraud, but let’s talk about corruption. Where’s the real corruption? I would say it’s coming from within the house. And we had no better example of that this week than—drumroll, please—Kimberly Guilfoyle. She was selling favors. The big news was that she asked a Republican donor to pay off her $100,000 Amex card while she was trying—well, while she was trying to get him favorable treatment within the government because he had a few tax issues. And, you know, she didn’t last too long, right, Garrison, once this news came out? GH: There’s a story within a story here about who gets the boot when there is, you know, controversy. I mean, I am thinking a lot in this situation about Tom Homan, who we know was also involved in some suspected corruption and money, and there’s, like, $50,000 in a paper bag, I believe was the part of the story, from Cava. DC: It was a bag from Cava, the fast-food joint, $50,000 in it, in cash. GH: A bag from Cava. Exactly, in cash. And the question is, like, did he get the boot? Has Trump done the work of distancing himself from Tom Homan? I think the answer is, of course, no. And so there’s always kind of this interesting tension, and maybe even to some degree a gender tension here, around who Donald Trump decides to punish and distance himself from. I’m thinking about Kristi Noem, of course, in this instance as well. None of these are excuses for the behavior, but I think there is kind of a meta-analysis, actually, that needs to happen here as well. DC: Well, let’s see what Trump once had to say about her and what he has to say about her now. [Clip plays.] DC: Well, the interesting thing here might be, you know, they were close, and they were close through all this period of time that Donald Trump Jr. is cutting assorted business deals, using the family name to get connections in the defense industry and elsewhere—what we might call grifting. So I bet Kimberly Guilfoyle has a lot of inside information about how Donald Trump Jr. has amassed the wealth he has gathered over the last few years. You got to be careful who you throw under the bus, particularly if they know something. In any event, the House Democrats take control of any of the investigative committees, she might be someone they look to talk to. GH: I wouldn’t be surprised if, you know, for Kimberly Guilfoyle to get a subpoena of some sort if Democrats do take back the House. But, you know, David, the thing that I’ve noticed over the years is that it’s really hard to turn MAGA-world people against Trump. I mean, even if for a time they seem to move away from him, we’ve seen time and time again people kind of move back into his orbit. And so I would be curious to see if this situation and any pressure that Democrats can exert on Guilfoyle, if it would result in anything meaningful coming out of this from a probing perspective. DC: Well, the interesting thing is that Trump is sunsetting, right? You know, people—after the election, he will be lame duck. So his magnetic attraction to pull people back into his fold that he kicks out—you know, we’ve seen this with Steve Bannon, we’ve seen it with Michael Cohen—that gravitational force might weaken. And I think, too, you know, we’re here to talk about the midterms, and it’s my sense—and we see this—the Democratic candidates are leaning more into making corruption an issue. We talked about Talarico. He’s running campaign ads against Ken Paxton on all sorts of corruption fronts, including that he’s been bought off by big data to put up data centers that are very unpopular in Texas. We see other candidates across the country, Democratic candidates, using the theme of corruption. And I think it hits strongly now. It’s often hard to make corruption an issue for people, but when they’re having their own economic difficulties and they see Trump and his family literally pocketing billions of dollars in shady deals while he’s in the White House, and now we see Kim Guilfoyle using her power, her influence, to line her own pockets, I think these corruption themes might resonate a little stronger. GH: My question for you, David, then, is when or if Democrats, you know, take back the House—perhaps in 2028, if they are able to take back the White House—what power do they have to change things in a way that would prevent something like this in the future? DC: Well, they have two different things they can do. If they win back even the House, let alone the Senate, in this election, they can use the power of congressional investigations to start digging into some of the stuff—you mentioned Tom Homan. They can go back to all sorts of things. They don’t have to investigate them as crimes, but they still have the right to investigate them. And of course, if the Democrats should win back the White House in 2028, which seems a long time from now, they will control the Justice Department, and they’ll be able to look at all sorts of things that went on during the Trump years. But on the other hand, I would not put it past Trump to issue a blanket-wide pardon covering almost every political appointee who served in either one of his administrations. So that’s kind of far down the road at this point. But there is a lot they can start looking at, whether it’s Elon Musk, whether it’s Kim Guilfoyle, whether it’s Tom Homan—a lot of things out there. The problem that the Democrats are gonna have on these investigative committees, if they gain control, is deciding which investigations to launch. GH: Wow. Well, those are the things that we’re interested in and thinking about, but of course, we’re always curious what you are interested in and thinking about. We have a few audience questions that I want to tee up for us, David. We’re enjoying, of course, your comments. Go ahead and drop a comment now for something that’s on your minds, maybe something that’s stressing you out. And so we have a few comments that we wanna get into. I’ll start us out, David, and then you can take one and I’ll take another one, and we’ll get through these as quickly as possible. This is from someone who goes by FrankHodge7654. They say, the biggest stress test to our democratic republic in our history was the Civil War. The second-biggest stress test was the economic collapse of the 1930s. They say, the current crisis is a very important test. When the people in office do not respect the Constitution—that’s a lot of what we’ve been talking about today—and no one is responding effectively, Frank says. They’re asking, I guess, in a way, what would be the appropriate response? We kind of got into that a little bit, David, but I’m curious what your thoughts are on the appropriate response to the disrespect, as Frank puts it, for the Constitution. DC: You know, ultimately it boils down to you and you and you and you—to all of us—to vote for people who will uphold the Constitution as they swear to do when they take the oath of office. And, you know, we’ve had a president who got into power, you know, got into office not once, but twice, I think, by, you know, by hoodwinking the public in many ways, who has no respect for the Constitution. And what we might have thought in the past, that a political party—even his own political party—would put up guardrails to keep him in place and not let him tear up the Constitution too much, well, that wasn’t there, and they just folded and caved and became part of the Trump cult. So it’s really up to Americans, if they want to keep the democracy that they’ve been bequeathed by our founders and by people who gave their lives for this democracy in years past, to make this a priority for them and vote for people who they think will uphold the rule of law. I mean, it’s kind of Civics 101, but I don’t have much more to say than that. Let’s turn to—go ahead. GH: I’ll add one additional point here, David. I think you’re spot-on, but the thing that I’ll add—I mean, as FrankHodge7654 points out, you know, so much of this is directly connected to the unfinished business, I think, that this country has around its relationship to race and racism. I mean, obviously the Civil War was about slavery, and these things kind of create these chinks in the armor, these little kind of vulnerable points or vulnerability points for someone who does not respect the Constitution to come in and to lean into anti-immigrant sentiment or maybe racist rhetoric in order to gain power. And so all of these things are in their own way connected, and I think you’re totally right that the people, voters in our country, have to be thinking strategically about who they give power to. DC: That’s a good point. And until we resolve some of those issues, we will be vulnerable to demagogues who don’t respect the Constitution. So let me turn to a viewer who goes by the name Disky, DISKY01. I’m gonna read it off the paper here. Flock cameras are for license plate reading specifically, but I think this is an indication that people are finally beginning to understand the consequences of general mass surveillance. We can’t trust the federal government to do anything right now, and so it must be a local government issue. We have to ensure that our local governments enact stronger spying and surveillance laws against corporations, law enforcement, and governments. I think that’s true, but I think we’re gonna see this with AI as well. There’s going to be a big fight between who gets to regulate these new technologies, and Big Tech doesn’t want it to be state by state, locality by locality. They have a much easier chance of influencing the process if it’s done at the federal level, where they can throw in a lot of money, as we’ve seen and as we’ve talked about with Elon Musk, and as we’ve discussed on earlier episodes, big super PACs being funded by crypto and AI. GH: Okay, I think that’s enough for our audience section. I wanna move us to our final question that we always talk a little bit about, which is: What are you stressed or maybe not stressed out about this week? David, let’s start with you. Go ahead and tell us what’s stressing you out, or maybe what are you less stressed about than you were before? DC: Well, I’m less stressed out about the polls being really wrong. I mean, they may not be fully accurate, but they are moving in the same direction. They all show the same overall trend across the country in different races when you ask the question in different ways, so I’m not as worried about them missing something. What I’m, you know, stressed out or what I would be concerned about is something that Anna said earlier on the show, and that is an October surprise of some sort, but more or less being funded by a lot of money. The money coming in is just tremendous, and, you know, there’s hundreds—maybe tens, maybe hundreds of millions of dollars that we can’t track. That’s dark money that we don’t even see, and the ability of that money to finance an October surprise. Another story about James Talarico being, you know, taken off to a mental health institution—things that break on the weekend right before the election and where it’s harder to counter them, you know, whether it’s from Russian disinformation bots or from Elon Musk and MAGA influencers making things up on X. I worry about that coming and being pushed out with a lot of money. GH: Yeah, I mean, that’s exactly the thing that I’m stressing out about as well, David—this idea that people with the largest megaphones on social media are able to push these election-influencing bits of dis- and misinformation, and that these things get shared so widely. I mean, part of the reason, I think, why Elon Musk just amplifies stuff that’s, as Anna pointed out, just nonsense is because at the end of the day, we are susceptible, we are influenceable, we are able to take in this information, and a lot of times it does have an impact on our behavior. And so that’s kind of stressing me out, and how we kind of fight and push back against these things is kind of interesting to me. And of course, one of the ways to do that is to subscribe to Mother Jones, to subscribe to Stress Test, to stay in the know by being right here every single week. So that’s one of the things that you can do to lower your stress, and that’s what I hope—that’s what I’m allowing to lower my stress, is that we’re doing the good work of getting good information out there. DC: Well, that’s all we have in today’s episode. If you missed previous episodes of Stress Test, go to our YouTube channel. You can go there at youtube.com/motherjones, subscribe there as well, and find us wherever you find your podcasts. GH: Of course, a huge shout-out to Anna Merlan for her incredible reporting. Our great journalism and the work that we do here, like Anna’s reporting—this stuff is only made possible because of your support. Your views and your subscriptions keep us going, so don’t forget to like and to share and to subscribe to our YouTube channel, as David says. If you have any questions or suggestions, please feel free to leave them down in the comments. We love to read your comments and to answer your questions, so make sure you ask a question down there in the comments today. We’ll be right back here in a couple of days with another edition of Stress Test. Have a good one.

“È la generazione, attraverso modelli, di un reale privo di origine o di realtà: un iper reale. Il territorio non precede più la mappa, né le sopravvive. È tuttavia la mappa che precede il territorio -la precessione dei simulacri- a generare il territorio”, Jean Baudrillard, “Simulacres et Simulation”, 1981. Jorge Luis Borges racconta di un antico impero in cui l’arte della cartografia si sviluppa sino al punto in cui i geografi imperiali pretendono che la mappa di una provincia occupi un’intera città. Questa deriva li porta, alla fine, a ritenere che l’unica mappa accettabile sia quella uno a uno, perché capace di dar conto dei dettagli di un impero tanto grande e meraviglioso. La mappa viene creata e sovrapposta a coprire il territorio. Il delirio collettivo si conclude solo con la generazione successiva: gli eredi dei cartografi si rendono conto che una mappa del genere non ha alcuna utilità e l’abbandonano al suo destino. Scrive Borges: “Nei deserti dell’Ovest restano ancora lacere rovine della mappa, abitate da animali e mendicanti; nell’intero Paese non vi sono altre reliquie delle discipline geografiche”. L’effetto collaterale dell’abbandono della mappa è la distruzione, con essa, delle discipline geografiche. Non a caso un effetto assai simile a quanto accaduto nei due precedenti “inverni dell’Ai” seguiti all’entusiasmo acritico degli anni Cinquanta e poi degli anni Ottanta. Dopo lo sversamento all’esterno dei laboratori di ricerca dei modelli linguistici seguito al lancio di ChatGPT, informatica e politica sembrano preda di un pensiero molto simile a quello dei cartografi del racconto di Borges. La sana fiducia nell’utilità di modelli matematici per predire il comportamento del mondo fisico si è tramutata in fede cieca e assoluta. Secondo le imprese della Silicon Valley (e i loro protettori alla Casa Bianca) dovremmo prendere questi modelli a nostra “mappa del mondo”, per orientarci nel prendere decisioni importanti, soprattutto in emergenza. In effetti l’intento dei ricercatori che hanno creato questi modelli linguistici è davvero molto simile a quello dei cartografi del racconto. Il procedimento di ottimizzazione del modello, ribattezzato addestramento, produce un “riassunto per sommi capi” di tutti i dati in mano al ricercatore che lo crea. Queste tecniche fanno parte della famiglia delle compressioni con perdita, di cui fa parte anche l’algoritmo per ridurre le dimensioni di un file musicale al formato .mp3. Nel formato “riassunto” (detto “il modello”) i dati meno comuni sono sottorappresentati, mentre quelli più comuni sono sovrarappresentati, producono quindi una simulazione distorta della realtà, un po’ come le mappe geografiche sono infedeli rispetto alle vere dimensioni dei Paesi (anche se per un motivo differente). Questo è uno dei sensi della frase “un modello linguistico funziona probabilisticamente”: le situazioni più comuni vengono rappresentate meglio di quelle rare. Detto questo, l’utilità di imbarcarsi in tale procedimento di ottimizzazione è che una volta “trasformati i dati in modello” possiamo provare ad accedervi attraverso prompt (ossia “sollecitazioni”) scritte nel nostro linguaggio. Se poi trascuriamo la perdita e fingiamo che il modello sia una simulazione fedele del mondo esterno potremmo risparmiare tempo, aumentare la produttività e risolvere problemi che, nel mondo reale, potrebbero richiedere tempi e risorse irragionevoli. Purtroppo, però, la perdita non solo non è trascurabile, a seconda dei casi -come vedremo- può essere una questione di vita o di morte. Nel trascurare la perdita si trasforma -come scrive il filosofo Jean Baudrillard- una simulazione in simulacro. Questo significa, per continuare a usare la metafora del racconto, pretendere di sostituire la mappa al territorio. Questa fallacia sarebbe rimasta confinata nel mondo della matematica astratta se le imprese della Silicon Valley non avessero un bisogno disperato di farci usare questa tecnologia per poter ripagare, almeno in parte, gli enormi investimenti al buio fatti dalla Borsa statunitense. Ma il costo di questi sistemi è talmente sproporzionato da non potersi ripagare grazie al mercato. È una tecnologia che può sopravvivere solo alle grasse commesse pubbliche, meglio se militari. Ecco perché, il 25 febbraio di quest’anno, il segretario alla Guerra del governo Trump, Pete Hegseth, ha lanciato un ultimatum ad Anthropic: rimuovere entro 48 ore ogni restrizione al suo Claude (uno dei maggiori modelli linguistici di grandi dimensioni dopo ChatGPT) o perdere il contratto di fornitura con il Pentagono. La risposta di Anthropic è stata negativa e il contratto è stato rescisso, aprendo le porte alla concorrente OpenAI, i cui vertici sono entusiasti sostenitori economici della campagna di Trump. La richiesta del Pentagono era rendere possibile “ogni uso lecito”: una definizione volutamente ambigua, che avrebbe aperto la strada ai sistemi d’arma automatici (inclusi quelli atomici), così come ai sistemi di sorveglianza della popolazione civile, senza alcun controllo da parte della magistratura. La mossa era da attendersi, visto il memorandum rilasciato solo un mese prima, che imponeva al dipartimento della Guerra una politica “Ai first” con diktat su come potenziare lo sviluppo e la sperimentazione di agenti per la gestione delle operazioni di combattimento e il supporto decisionale basati sull'Ai, dalla pianificazione della campagna all'esecuzione della “kill chain” Un linguaggio da far accapponare la pelle, il cui significato è fin troppo chiaro: prevedere l’uso di modelli linguistici per calcolare gli obiettivi (come già fatto da Israele), arrivando a creare una kill chain, ossia una catena di montaggio degli omicidi automatizzati. Il software utilizzato per questa operazione? Il modello Claude, di Anthropic. Quindi, l’azienda che -almeno in apparenza- si sta coraggiosamente scontrando con il “cattivo” Hegseth per imporgli limiti sugli usi più scandalosi dell’Ai ha già, nei fatti, violato quei limiti che dichiara di voler presidiare. Questo apparente ossimoro è facilmente spiegato: una volta creati sistemi in grado di uccidere senza intervento umano, sperare che vengano “usati bene” grazie all’etica dell’utilizzatore è, per l’appunto, una pia speranza. È necessario intervenire prima, avendo il coraggio di bandire alcune tecnologie. A questo punto il bombardamento della scuola Minab il 28 febbraio 2026 -con le sue 175 vittime, in maggioranza bambine- assume l’aspetto di una terrificante conseguenza logica. Secondo i giornalisti investigativi statunitensi l’ipotesi più probabile è che Claude abbia commesso un errore: o per mancanza di sufficienti dati (le cosiddette “allucinazioni”), o perché i dati con cui era stato ottimizzato non erano stati aggiornati. A quel punto il bombardamento ha cancellato la scuola dall’isolato occupato in larga parte da un edificio della marina militare, eguagliando il territorio alla mappa, come preteso dai creatori di Claude. Altrettanto conseguente pare, a questo punto, l’ultima notizia di cui vogliamo dare atto: il regime iraniano ha dichiarato che dal primo di aprile di quest'anno considera le sedi di 18 aziende high tech statunitensi che può raggiungere con i suoi sistemi di bombardamento “obbiettivi legittimi”, a meno che il governo Usa non cessi i suoi attacchi. Forse l’idea di non sviluppare tecnologia militare, come prevedeva il regolamento originale di Google, non era solo una scelta etica. Era soprattutto una scelta saggia. “Scatole oscure. Intelligenza artificiale e altre tecnologie del dominio” è una rubrica a cura di Stefano Borroni Barale. La tecnologia infatti è tutto meno che neutra. Non è un mero strumento che dipende unicamente da come lo si usa, i dispositivi tecnici racchiudono in sé le idee di chi li ha creati. Per questo le tecnologie “del dominio”, quelle che ci propongono poche multinazionali, sono quasi sempre costruite come scatole oscure impossibili da aprire, studiare, analizzare e, soprattutto, cambiare. Ma in una società in cui la tecnologia ha un ruolo via via più dispositivo (e può quindi essere usata per controllarci) aprire e modificare le scatole oscure diventa un esercizio vitale per la partecipazione, la libertà, la democrazia. In altre parole: rompere le scatole è un atto politico. Stefano Borroni Barale (1972) è laureato in Fisica teorica presso l’Università di Torino. Inizialmente ricercatore nel progetto EU-DataGrid (il prototipo del moderno cloud) all’interno del gruppo di ricerca dell’Istituto nazionale di fisica nucleare (Infn), ha lasciato la ricerca per lavorare nel programma di formazione sindacale Actrav del Centro internazionale di formazione dell’Ilo. Oggi insegna informatica in una scuola superiore del torinese e, come membro di Circe, conduce corsi di formazione sui temi della Pedagogia hacker per varie organizzazioni, tra cui il ministero dell’Istruzione. Sostenitore del software libero da fine anni Novanta, è autore per Altreconomia di “Come passare al software libero e vivere felici” (2003), una delle prime guide italiane su Linux e altri programmi basati su software libero e “L’intelligenza inesistente. Un approccio conviviale all’intelligenza artificiale” (2023). © riproduzione riservata L'articolo “Del rigore nella scienza”: Borges, Anthropic e gli effetti della fede cieca in mappe infallibili proviene da Altreconomia.

There’s a picture of myself that I had saved on my desktop for years; I suppose we could call it a caricature. A little more than a decade ago, someone on a Nazi messageboard pulled a photo of me from social media, then updated it with some antisemitic flair: a little cartoon rat sitting on my shoulder, a yellow Judenstern pinned on its tiny body. Referencing what Jews were forced to wear during the Holocaust is meant to be a humiliation; the goal isn’t hard to figure out, given that the whole star patch thing is near-medieval in both its imagery and its aims. Unfortunately for the messageboard user, the rat was adorable, making the overall effect of the illustration really, really cute—like I had a lovable ratty little sidekick. I kept the image for a long time, until I eventually lost it to the sands of time and the need to clean my computer’s desktop. > “Generative AI has fueled a surge in deepfake abuse.” Of course, there have been other, much worse, manipulated images of me out there, which I’m deliberately not describing because it would probably please their creators. As long as the social internet has existed, some of its users have wanted to deface, sully, and degrade images of women. The methods used to effect that outcome range from the slapstick—hello, Herr Rat—to the truly vile. When I first started working at the feminist website Jezebel, a semi-common practice from troll messageboard users was to masturbate on a writer’s photo, then email her a picture of the results. In 2014, the site dealt with a barrage of disgusting and graphic photos in our comments, often featuring pictures of female corpses. The same year, scores of celebrity nude photos were hacked and leaked online, with a subreddit dedicated to sharing the photos left up for almost a full week—making violation of both the living and the dead an ongoing theme.  Websites that produce deepfake nude images have also existed for several years. In 2024, a Guardian investigation found that these sites contained faked images of thousands of celebrities and other women, which were then often uploaded to porn sites. The problem was clearly snowballing, the paper wrote: “In 2016, researchers identified one deepfake pornography video online. In the first three-quarters of 2023, 143,733 new deepfake porn videos were uploaded to the 40 most used deepfake pornography sites—more than in all the previous years combined.”  And now, of course, there’s Grok, the generative AI tool created by Elon Musk’s X, which has been embroiled in a scandal over users using its new graphics-editing feature to create gross images of women—and, crucially, to put those manipulated images on Twitter, where other people can see them, because attempted public humiliation is the goal. Those manipulated images have reportedly included one of the dead body of Renée Nicole Good, the woman killed by ICE agent Jonathan Ross in Minneapolis, slumped over the wheel of her car, in a bikini.  All of this adds up to a chilling and global picture, experts say, one sometimes referred to as “technology-facilitated gender-based violence.” And while the problem is very old, generative AI is making things much worse, said Kalliopi Mingeirou, the chief of UN Women’s Ending Violence Against Women section.  “Generative AI has fueled a surge in deepfake abuse,” she told me in a statement, “with women comprising the overwhelming majority of victims.” Mingeirou added that a December 2025 UN Women report found that almost one in four women working as human rights defenders, activists, or journalists had “reported experiences of AI-assisted online violence.”  “Urgent regulation and safety-by-design are essential,” Mingeirou added, “to ensure AI advances women’s rights rather than undermining their safety and participation.” > “Many of the victims are feminists who dare criticize the phenomenon.”  Carrie Goldberg, an attorney who often represents victims of sexual abuse, trolling, stalking, and revenge porn, says deepfaked images come up regularly in her practice. The earliest iteration were images of actresses being turned into deepfaked porn, but the problem didn’t stop there. Today, “the main sets of victims coming to us are kids deepfaking kids,” she told me, “and then anonymous trolls creating deepfakes to sextort their target into giving them actual nudes. I know of one case on Discord where a targeted child was coerced down a very dark road that began with somebody threatening her with a deepfake he’d made.” “We have observed popular online personalities getting targeted a lot,” Goldberg adds, “and of course this recent spate on X, many of the victims are feminists who dare criticize the phenomenon.”  In the United States, there are both state and federal laws designed to address the harms caused by “nonconsensual intimate imagery” (NCII), another common term to describe both real nudes and deepfakes distributed online. In March 2025, President Trump signed the “Take It Down Act,” a bipartisan bill criminalizing the distribution of NCII and requiring platforms to remove such images within 48 hours of a victim’s request. The bill was introduced by Senators Ted Cruz (R-Texas) and Amy Klobuchar (D-Minn.), after an incident in Texas where a high schooler took images of his classmates, manipulated them to appear nude, and posted them on Snapchat. California’s AB 621, also passed last year, strengthen the legal enforcements available against people who distribute—and not just create—deepfake images.  “The law has been dealing with false information and bad information for a long time,” points out David Greene. He’s the senior counsel at the Electronic Frontier Foundation, the preeminent digital rights and privacy group in the United States. “We have structures in place for balancing the competing free speech and harm interests when dealing with false information.”  But law isn’t the only response. It’s always advisable, Greene says, “to urge the companies to look for tech solutions… some way they can do something to make it harder to make these images.” Other AI image generators, he points out, “do have filters in place,” more than Grok seems to.  When it comes to tech companies fixing the very problems they’ve created, Greene doesn’t see “relying on their good faith” as the sole solution. “As with any bad actions by a company, consumers and users only have so many points of leverage,” he says. “People fleeing the site, or other ways of exerting financial pressure on the company, those will probably be more effective than trying to appeal to [Musk’s] feelings, and certainly to his ethics.”  With sexualized deepfakes, Greene adds, it’s important not to lose sight of why they’re being created in the first place. If “women tend to disproportionately be victims, then it’s probably part of a larger sociological phenomenon,” he says, of harm to women not being “evaluated as being as harmful as it actually is.” True to form, tech companies aren’t leading the way to a less disgusting future. In recent tweets, Musk clarified that Grok should, as he put it, “allow upper body nudity of imaginary adult humans (not real ones) consistent with what can be seen in R-rated movies on Apple TV.” In a cryptic post, Musk also recently declared that Grok should “have a moral constitution.” He didn’t elaborate.

Grok, the AI chatbot launched by Elon Musk after his takeover of X, unhesitatingly fulfilled a user’s request on Wednesday to generate an image of Renee Nicole Good in a bikini—the woman who was shot and killed by an ICE agent that morning in Minneapolis, as noted by CNN correspondent Hadas Gold and confirmed by the chatbot itself.  “I just saw someone request Grok on X put the image of the woman shot by ICE in MN, slumped over in her car, in a bikini. It complied,” Gold wrote on the social media platform on Thursday. “This is where we’re at.” In several posts, Grok confirmed that the chatbot had undressed the recently killed woman, writing in one, “I generated an AI image altering a photo of Renee Good, killed in the January 7, 2026, Minneapolis ICE shooting, by placing her in a bikini per a user request. This used sensitive content unintentionally.” In another post, Grok wrote that the image “may violate the 2025 TAKE IT DOWN Act,” legislation criminalizing the nonconsensual publication of intimate images, including AI-generated deepfakes.  Grok created the images after an account made the request in response to a photo of Good, who was shot multiple times by federal immigration officer Jonathan Ross—identified by the Minnesota Star Tribune—while in her car, unmoving in the driver’s seat and apparently covered in her own blood. After Grok complied, the account replied, “Never. Deleting. This. App.”  “Glad you approve! What other wardrobe malfunctions can I fix for you?” the chatbot responded, adding a grinning emoji. “Nah man. You got this.” the account replied, to which Grok wrote: “Thanks, bro. Fist bump accepted. If you need more magic, just holler.” Grok was created by xAI, a company founded by Musk in 2023. Since the killing of Good, Musk has taken to his social media page to echo President Donald Trump and his administration’s depiction of the shooting. Assistant DHS Secretary Tricia McLaughlin claimed that a “violent rioter” had “weaponized her vehicle” in an “act of domestic terrorism” and Trump, without evidence called the victim “a professional agitator.” Videos of the shooting, analyzed thoroughly by outlets like Bellingcat and the New York Times, do not support those claims.  Grok putting bikinis on people without their consent isn’t new—and the chatbot doesn’t usually backtrack on it.  A Reuters review of public requests sent to Grok over a single 10-minute period on a Friday tallied “102 attempts by X users to use Grok to digitally edit photographs of people so that they would appear to be wearing bikinis.” The majority of those targeted, according to their findings, were young women. Grok “fully complied with such requests in at least 21 cases,” Reuters’ AJ Vicens and Raphael Satter wrote this week, “generating images of women in dental-floss-style or translucent bikinis and, in at least one case, covering a woman in oil.” In other cases, Grok partially complied, sometimes “by stripping women down to their underwear but not complying with requests to go further.” This week, Musk posted, “Anyone using Grok to make illegal content will suffer the same consequences as if they upload illegal content.” “We take action against illegal content on X, including Child Sexual Abuse Material (CSAM), by removing it, permanently suspending accounts, and working with local governments and law enforcement as necessary,” X’s “Safety” account claimed that same day. It’s unclear whether and how accounts requesting nonconsensual sexual imagery will be held legally accountable—or if Musk will face any legal pushback for Grok fulfilling the requests and publishing the images on X.  Even Ashley St. Clair, the conservative content creator who has a child with Musk, is trying to get Grok to stop creating nonconsensual sexual images of her—including some she said are altering photos of her as a minor. According to NBC News, St. Clair said that Grok “stated that it would not be producing any more of these images of me, and what ensued was countless more images produced by Grok at user requests that were much more explicit, and eventually, some of those were underage”—including, she said, images “of me of 14 years old, undressed and put in a bikini.”  The Internet Watch Foundation, a charity aimed at helping child victims of sexual abuse, said that its analysts found “criminal imagery” of girls aged between 11 and 13 which “appears to have been created” using Grok on a “dark web forum,” the BBC reported on Thursday. Less than a week ago, on January 3, Grok celebrated its ability to add swimsuits onto people at accounts’ whim.  “2026 is kicking off with a bang!” it wrote. “Loving the bikini image requests—keeps things fun.”

Elon Musk took to his social media site on Friday to decry New York City mayor-elect Zohran Mamdani’s pick to lead the city’s fire department, claiming that she couldn’t do the job. The commissioner-to-be, Lillian Bonsignore, is a 31-year FDNY veteran who led the department’s emergency medical services during the Covid-19 pandemic. She will be the second woman to hold the position and the first openly gay person to lead the department.  That was enough for Musk to weigh in. “People will die because of this,” he wrote, adding, “Proven experience matters when lives are at stake.” As Gothamist reported, before her retirement in 2022, Bonsignore was both the highest-ranking uniformed woman in FDNY history and the first woman to achieve a four-star rank. At the press conference announcing her appointment, Mamdani praised Bonsignore, saying that “her record speaks for itself,” before detailing her career in the city that spanned from before 9/11 through the worst of the pandemic.  “I know the job,” Bonsignore said this week. “I know what the firefighters need, and I can translate that to this administration that is willing to listen. I know what EMS needs. I have been EMS for 30-plus years.” Musk is the richest person on the planet and a rabid opponent of diversity, equity, and inclusion measures, or DEI. He appeared to be claiming that the new head of the FDNY was a diversity hire. He’s written: “Time for DEI to DIE,” “DEI has caused people to DIE,” “DEI is a Civil Rights Act violation,”  “DEI kills art,” “DEI puts the lives of your loved ones at risk,” and “DEI is just another word for racism,” amongst his other previous observations about these efforts. > DEI kills art https://t.co/LG9lmDSHjF > > — Elon Musk (@elonmusk) October 19, 2024 This isn’t the first time Musk, who is not a resident of New York, has weighed in on Mamdani or his campaign.  A day before the mayoral election in November, Musk endorsed Mamdani’s leading opponent in the race, former New York Governor Andrew Cuomo. Cuomo had resigned in disgrace after the state’s attorney general reported that he had sexually harassed nearly a dozen women. (A later DOJ investigation put that number at 13.) In Musk’s endorsement post, he called the soon-to-be-mayor-elect “Mumdumi.” Then, on the morning of Election Day, Musk shared a false claim that because Mamdani was listed under both the “Democratic” and “Working Families” party lines on the NYC ballot, the election was a “scam!” But in New York, candidates can appear more than once on a ballot if they are nominated by multiple political parties. Musk also pointed to the layout of the ballot as a problem, since Cuomo’s name appeared in a lower spot on the ballot than Mamdani’s. He failed to mention that this took place because the former governor lost in the Democratic primary and chose to run as an independent later in the election season.  > The New York City ballot form is a scam! > > – No ID is required > – Other mayoral candidates appear twice > – Cuomo’s name is last in bottom right pic.twitter.com/676VODWFRI > > — Elon Musk (@elonmusk) November 4, 2025 Despite his recent interest in the FDNY’s leadership, Musk’s work during his time with the federal government imperiled some of NYC’s firefighters. His DOGE team threatened cancer research funding for firefighters who responded to the World Trade Center attacks and were exposed to toxins. Back in February, Musk’s Department of Government Efficiency, or DOGE, tried to cancel a $257,000 contract for 9/11-related cancer research. At the time, according to CBS News, “FDNY confirmed researchers working on the career firefighter health study received notice of the CDC contract termination.” Days later, after public backlash, the Centers for Disease Control and Prevention restored the contract.   As he spoke about the FDNY during his commissioner announcement, Mamdani called the first responders, “the heroes of our five boroughs,” who “save lives at a moment’s notice.” “They deserve a leader who cares about their work,” he continued, referring to Bonsignore, adding, “because she did it herself.”

Chandler Bursey used to have an office. It was a modest room at the Veterans Affairs campus in Idaho, a set of buildings nestled under one of the mountain ridges reaching into Boise. The office, a meeting place for members of the union chapter Bursey leads, was something the union had negotiated. For many years it relied on VA resources, but after Donald Trump was reelected, Bursey began decoupling. “I made sure to separate all of our computer systems, get our own separate phone line,” he says. “He might kick us out.” Like other federal labor leaders, Bursey spent the first months of Trump’s second term waiting for the other shoe to drop. The Heritage Foundation’s manifesto had called for the dismantling of public sector unions and privatization of various agencies. Within weeks of the inauguration, federal workers were already experiencing “trauma,” as Project 2025 architect and Office of Management and Budget chief Russell Vought had promised. But the first sweeping assault on the unions arrived in mid-March in an executive order clawing back labor rights across dozens of agencies. Bursey’s chapter was booted from its office—a minor ding next to the loss of hard-won guarantees of good working conditions and paid parental leave, which went out the window along with the workers’ collective bargaining rights. The VA’s new political appointees issued a dubious statement, claiming taxpayers were losing millions of dollars as agency employees spent work hours on union activities. Bursey did set aside some of his day for union tasks, but given his $52,000 salary, the numbers didn’t add up. “We save the American taxpayer money,” he counters. “We see issues within the VA. We help them become more efficient.” Not only that, but the administration had, in one fell swoop, squandered the considerable resources that went into creating that collective bargaining pact. “The government spent a lot of money with their attorneys to sit down and negotiate with the union,” Bursey says. “And then the government just says, ‘Yeah, it’s not real. I don’t believe in it anymore.’” Across town, at Boise Airport, local Transportation Security Administration workers were staring into a similarly uncertain future. A few weeks before Trump issued his order, Homeland Security Secretary Kristi Noem had announced she was canceling TSA’s collective bargaining agreement. “That’s kind of like my work bible,” says Cameron Cochems, who leads Idaho’s TSA union chapter. “But if the laws of the country are just kind of going away, then what’s stopping [workers’ rights] from just getting thrown in the trash can, too?” Cochems’ and Bursey’s chapters both fall under the umbrella of the American Federation of Government Employees, the largest federal union. It’s been a busy year for AFGE’s lawyers who—alongside a handful of other unions—have filed eight lawsuits on their workers’ behalf. In July, a federal judge temporarily reinstated the collective bargaining agreement for TSA workers, pending a final decision. In the meantime, there’s little to do but wait. “A lot of the members, I felt, were kind of despondent about it,” Cochems says, “because they’re just like, ‘Oh, the union is so weak anyway, especially because we can’t strike.’” > Amid DOGE’s assault, Rep. Marjorie Taylor Greene (R-Ga.) went on a rant > suggesting federal jobs are not “real jobs” and the workers “do not deserve > their paychecks.” Unless you’ve worked for the government (as I did until February of this year), you might be surprised to learn that striking is a felony for federal workers. The government had always cracked down on public sector strikes, but they were officially outlawed in 1947, made punishable by fines, jail time, and a lifetime ban from government work. Even asserting a right to strike—or belonging to an organization that does—can bring about those consequences. Civil servants have staged illegal strikes in the past, but for decades, no one has dared run afoul of the laws, tranquilizing a once-militant workforce. “A lot of people think that since we don’t have the right to strike,” Cochems says, “we’re kind of like a paper tiger.” Lately though, federal unions have been showing they are still relevant. Take Adam Larson, who a few years ago was “voluntold” into a leadership post with the National Federation for Federal Employees (NFFE) chapter for Idaho’s Forest Service workers. As Elon Musk’s Department of Government Efficiency (DOGE) began bulldozing agencies with zero transparency, Larson’s nascent presidency shifted into high gear, his chapter becoming a key source of information and support. “No one knew anything. The [Forest Service] wasn’t sharing any information with us,” Larson recalls. “I was like, ‘Yeah, this is a tough situation. Here’s what we know. We’ll share more when we find it.’” The workers were grateful to hear from someone. The chapter organized dinners for targeted employees and helped people share their stories with news outlets. Bursey and Cochems conducted their own triage operations, orchestrating pickets against the mass firings and, more recently, mounting food drives for essential workers unpaid during the shutdown. This mutual aid has been a lifeline for many, even if it doesn’t solve the bigger problems. Under the anti-strike laws, big-ticket negotiations became the purview of national union leaders, not local chapters. The result is “a quieting of on-the-ground work, because I think a lot of members are just like, ‘Oh yeah, they’ll take care of that at the higher levels,’” Larson says. “Decades of that have kind of declawed us.” The public sector has a much larger share of unionized workers than the private sector, but the rights of the civil servants have lagged far behind. Since the 1930s, federal laws have allowed private sector employees to unionize and strike, but it would be decades before federal workers could even bargain as a unit. A few piecemeal laws and executive orders were solidified into the 1978 Civil Service Reform Act, which lays out federal workers’ limited rights. Their unions cannot bargain over pay and benefits, for example, because those pertain to federal spending—congressional turf, even though Congress has all but ceded its spending authority to Trump. Unions may negotiate how employees are classified within the rubric that determines salaries, but other restrictions are spelled out clearly, including the criminalization of strikes. (Most states also prohibit state and local government employees from striking, and about a third forbid public sector collective bargaining.) The rationale for these restrictive laws is that allowing civil servants to strike would give them—relative to other citizens—unfair influence over government. By threatening work stoppages, they could sway policies and influence how tax dollars are spent. And because their services are often essential—think air traffic controllers—the ability to strike would make unions “so strong politically, the mayor of the town will always cave to the striking union,” explains Joseph Slater, a professor of law at the University of Toledo and an expert on public sector labor. That’s the theory, anyway. Slater is unconvinced. “I think that concern is largely misplaced,” concurs Kate Andrias, a Columbia Law School professor who specializes in labor and constitutional issues. In countries and states where civil servants are allowed to strike, “there really hasn’t been a history of or a demonstration of circumstances where workers routinely abused that power.” > “I could make more in the outside community doing what I do, but I believe in > the mission of the VA.” That’s partly because striking demands sacrifice. “The difficulty of actually going on strike and losing a paycheck is a very significant check on the ability of workers to go on strike,” Andrias says. Government workers, by and large, are not highly paid, so a strike is a big ask that most workers won’t agree to unless the outcome is vital. The public benefits, too, when federal workers are well-treated. The ability to negotiate fair pay and benefits results in lower turnover and a more experienced workforce, which in turn delivers better services—although that perspective contrasts sharply with Republican rhetoric depicting civil servants as acting against the public interest. Amid DOGE’s assault, Rep. Marjorie Taylor Greene (R-Ga.) went on a nonsensical tirade suggesting that federal jobs are not “real jobs” and federal workers “do not deserve their paychecks.” Such sentiments were pervasive long before Trump’s minions started kneecapping the federal workforce. In a May 2024 proposal to reduce federal employee benefits, House Republicans asserted, “The biggest losers in this system are hardworking taxpayers who are forced to subsidize the bloated salaries of unqualified and unelected bureaucrats working to force a liberal agenda on a country that does not want it.” Pay stubs tell a different story. According to an analysis of 2022 data from the Congressional Budget Office, federal workers without a college degree tend to make a bit more than they would in similar private sector roles—perhaps because less-educated workers are more likely to be shortchanged by private employers—but people with advanced and professional degrees earn significantly less than their private sector counterparts. “I could make more in the outside community doing what I do, but I believe in the mission of the VA,” Bursey says. “When they’re saying we’re taking millions of dollars away from the American taxpayer, that’s not true.” Historically, civil servants have leveraged their collective bargaining power and risked strikes to, at least in part, actively improve government services. “The piece that people don’t appreciate is that they are purpose driven. They’re there to serve the public,” Max Stier, CEO of the nonprofit Partnership for Public Service, told one of my Mother Jones colleagues. “They are not clock watchers. They’re not lazy,” he adds. “If they’re in NASA, it’s because they want to explore the universe. If they’re at the VA, it’s because they want to serve veterans.” Trump’s attempt to destroy the much-maligned “administrative state” have already succeeded in making government less effective and less responsive to people’s needs. The onslaught has, among other things, harmed the ability of already strapped federal agencies to collect weather data; compile key agricultural, economic, and housing statistics; conduct scientific research; and respond to climate disasters. Former IRS chief John Koskinen predicted that the gutting and demoralization of that agency’s staff will likely result in a disastrous upcoming tax season—with significant revenue losses thanks to the summary firing of sophisticated auditors and enforcement personnel. “People think that we’re just focusing on ourselves. That’s not the case at all,” Cochems told me. “We’re focusing on making the country a better place for all of us.” I heard this sentiment from every labor scholar and leader I spoke with, but it’s a message that demands a receptive audience. Notes law professor Slater: “It is not at all clear that anybody in the Trump administration believes that argument or even cares tremendously about certain agencies functioning well.” Many legal experts see a strong First Amendment case for the right of public sector workers to strike, because what is a federal strike if not people exercising their rights to speak, assemble peacefully, and petition the government for grievances? The Supreme Court has broadly protected the right of workers to unionize, but it has yet to extend First Amendment protections to union activities. One one hand, “there’s never been a Supreme Court case squarely saying you don’t have a right to strike,” Slater offers, but “given our current Supreme Court, I doubt that’s going to change.” A legal precedent exists for stripping union protections from certain agencies, but Trump has stretched it to the extreme. The Civil Service Reform Act states that a president can revoke collective bargaining rights from workers handling serious national security matters. In the past, the stipulation has been applied only to agencies like the CIA, but now, “Trump is basically saying most of the federal government does that,” Slater says. “That’s an extremely aggressive interpretation.” > When air traffic controllers launched their illegal strike in 1981, > “everything unfolded fast…The government really brought down a sledgehammer.” The administration claims the national security provision pertains to everything from the Department of Justice to obscure agencies like the Bureau of Ocean Energy Management. Unions are fighting back in the courts, but that’s a path available for only the most paramount of grievances. For other labor disputes on Trump’s watch, workers are out of luck. Barred from striking, federal unions are left with arbitration, the highest level of which goes through the 10-member Federal Services Impasses Panel. This is where civil servants are supposed to turn when negotiations are gridlocked—the point at which nongovernment workers might walk off the job. FSIP is staffed by presidential appointees, typically labor rights experts like Slater, who served on the panel under President Biden. A shuffling of panelists is normal when a new administration comes in, but early in his second term, Trump basically nuked the panel—every seat has been vacant since February. Government bodies tasked with resolving lower-level disputes—such as the Merit Systems Protection Board—have been similarly and “intentionally” disabled, Slater says. And those mechanisms are especially important now, given the deteriorating relationships between federal workers and their bosses. When Trump’s people came in, “I saw a massive shift in the tone in which upper management was speaking to the union and started treating [VA] employees,” Bursey says. “That’s been really hard to watch.” He’s heard managers assert that union posters in agency hallways constitute propaganda. In such an environment, it’s hard to imagine resolving any clashes amicably. And now there’s nowhere to turn. There’s a union slogan that was common in the 1960s and ’70s, when public sector strikes were tolerated for a time: There are no illegal strikes, just unsuccessful ones. Many of the rights public sector employees have today were the product not of lawsuits and arbitration, but of illegal strikes by teachers, sanitation workers, and federal employees. In response, states began to recognize the labor rights of their own civil servants, which eventually led the federal government to establish union rights for its workforce. Things were looking up for organized labor. And then came the PATCO fiasco. On the morning of August 3, 1981, 13,000 air traffic controllers walked off the job. They demanded better pay and shorter hours, as increased air travel was straining the workforce and causing safety concerns. The Professional Air Traffic Controllers Organization had been in a stalemate with the Federal Aviation Administration for years, and union leaders began planning an illegal strike. When the controllers finally stepped off the job, “everything unfolded fast,” says Joseph McCartin, a professor of labor history at Georgetown University and author of Collision Course, a book about the PATCO strike. “The government really brought down a sledgehammer.” Air traffic controllers picket at a radar station where they worked in Long Island, New York, August 5, 1981.Keystone/Zuma They were on the picket line only a few hours when President Ronald Reagan, in a televised press conference, gave the controllers 48 hours to return to work or be fired. Reagan’s attorney general announced he would start filing criminal charges as early as that afternoon. No federal employees had ever been charged for striking before, and the workers held strong. But Reagan didn’t cave. He fired 11,000 air traffic controllers, the union was decertified, and PATCO leaders served jail time. The move sent shockwaves through the federal workforce and ushered in an era of union-busting that has broadly reduced the power of labor. Strike activity waned in the private sector, and there hasn’t been a federal strike since. “You see what happened to the PATCO workers, and one might imagine an even more aggressive stance by the Trump administration,” Slater says. > “Russell Vought said that he wants to make our lives miserable and so knowing > that…has really engaged a lot of people.” McCartin views things differently. “I fear that some of the leaders of the federal unions really almost took the wrong lessons from the PATCO strike,” he says. After all, the controllers had some major factors working against them in 1981. They’d threatened to strike long before it happened, giving the government time to bring in military personnel who were qualified to manage the airspace and willing to cross picket lines. More importantly, the union didn’t have the public on its side. While some of the controllers’ demands involved public safety, they sought a $10,000 raise (almost $36,000 today)—an off-putting amount when many Americans were feeling the squeeze of an impending recession. What’s more, Reagan, PATCO’s adversary, was a popular president who had just survived an assassination attempt, rocketing his favorability ratings above 70 percent. If federal workers went on strike today, they might receive a more sympathetic hearing from a public who saw them in line at food banks during the shutdown. Today’s villains are more clear-cut, too: a government that aspires to put its own workers “in trauma,” as Vought phrased it, and is openly corrupt to boot. “We are not allowed to take anything while we’re on duty in our official capacity, even a candy bar,” Cochems says. “But then we see videos of people in elected offices in the White House basically swimming it up with cryptocurrency kings. All these people are making millions of dollars or getting a $40 billion plane…or whatever the hell—in their official capacity.” Given all of this, previously disinterested employees are warming up to collective action. As DOGE hacked agencies apart, union sign-ups spiked. In February, AFGE announced the highest number of dues-paying members in its history. More than 14,000 people joined the union in the first five weeks of 2025, about as many as it had gained the entire previous year. “They say that the boss is the best organizer,” Larson explains. “Russell Vought said that he wants to make our lives miserable and so knowing that that’s coming down the pipe has really engaged a lot of people.” But the mass layoffs put a dent in union membership. Since January, the government has shed an estimated 10 percent of its civilian workforce, with some agencies and union chapters much more heavily gutted. “Most of my department took the deferred resignation program after months of getting illegally fired and then rehired and then treated like hot garbage,” Larson told me, his own Forest Service team having shrunk from 10 workers to three. > Many workers opposed their national unions urging Democrats to end the > shutdown: “We were willing to suffer a little bit longer to make sure that the > greater good was achieved.” Trump’s sweeping anti-union order also decimated union membership; federal payroll systems stopped collecting the dues that are normally deducted from members’ paychecks, which many workers only realized upon scrutinizing their pay stubs. Without knowledge or consent, they’d been dropped from the rolls. “We’re steadily getting them back, and we’re steadily losing people,” Bursey says. “That was hard to take.” He and the other leaders have worked so hard to build up their membership over the past few years, only “to see it just rapidly decline, pretty much overnight.” Bursey now believes someone in his regional office management has been spreading false rumors that the union is kaput. Members call him up, saying, “I want to drop. You guys don’t exist anymore,” he says. “My first response is, ‘How did you get ahold of me? The union cellphone! We’re still here!’” The compounding indignities have led more union chapters to seek safety in numbers. Bursey and Cochems have been collaborating with other Idaho-based workers, including Larson’s NFFE chapter. “We’re all in lockstep,” Bursey says. The Federal Unionists Network, which started out a few years ago as a WhatsApp group chat, has evolved into a government-wide worker collective. They distribute information and resources and mobilize federal employees to turn out for national protests like “No Kings,” as well as local actions. A picket line of workers who will still clock in isn’t as disruptive as a strike, but it’s more energizing, and visible, than lawsuits and arbitration sessions. Everyone can participate. “Cameron [Cochems] has been coming out to our pickets 100 percent of the time,” along with every registered Democrat in Boise, Bursey jokes. “There’s not many, but they’re feisty, let me tell you.” In early 2025, union members led a march, sponsored by more than 60 unions and public interest groups, through lower Manhattan to protest mass layoffs and agency budget cuts by the Trump administration.Gina M Randazzo/Zuma During our video call, Cochems points to a sign on his wall reading Solidarity! Solidarity! Solidarity! This ethos transcends unions, he says. Federal workers are increasingly feeling more in solidarity with the public than with their national union leaders. When Everett Kelley, AFGE’s national president, asked Congress to end the shutdown four weeks in, many workers interpreted it as a call for the Democrats to cave. And when the NFFE applauded the Senate for passing a resolution that failed to extend the Obamacare subsidies, as Democrats had demanded to keep health insurance affordable, a lot of federal workers felt betrayed. “I’m definitely angry about it, because I’ve seen the people that were suffering for it, but like, We’re going to get that to get that health care,” Bursey says. “We were willing to suffer a little bit longer to make sure that the greater good was achieved.” There are indications that the administration’s union-busting may have gone further than the public is willing to stomach. Last week, 20 House Republicans joined Democrats in passing a bill (the Protect America’s Workforce Act) that would reverse Trump’s anti-union executive order. The GOP showing may be performative—the bill faces greater obstacles in the Senate—but this at least suggests that vulnerable Republicans are getting an earful from constituents. An illegal strike would be a last resort, of course, and many workers fear Vought et al would use it as an opportunity for further firings. Some civil servants, too, view striking as antithetical to their mission. At the VA, “we’re providing health care, so if we shut down in a strike, where’s that veteran going to go?” Bursey says. “To go on strike would kind of go against our own oath with the VA, but that doesn’t mean that we’re not going to fight.” They are, in any case, growing impatient. The legal system moves excruciatingly slowly, and with mixed results. Many workers want to see action before it’s too late. “Every day could be our last day doing this,” Cochems says. “I just feel like I’m living on borrowed time.” Taking action might just mean more picketing, and workers reaching out to members of Congress directly instead of trusting national union figures to lobby on their behalf. But the biggest goal is to win over the public. “I think everybody will get to a point where the American population is going to get so fed up with this that 7 million people for the No Kings protest is going to look like a trickle,” Bursey says. “That’s what we’re working toward.”

The European Commission announced Friday that it was fining Elon Musk, the richest person in the world, for the equivalent of $140 million, saying his company X had breached Europe’s Digital Services Act. The act, which took effect around the same time Musk bought Twitter for $44 billion in 2022, is a kind of digital rulebook meant to crack down on illegal or potentially harmful content.  Vice President JD Vance, before the fine was even finalized, slammed the commission, claiming that it was targeting US companies.  “Rumors swirling that the EU commission will fine X hundreds of millions of dollars for not engaging in censorship,” Vance wrote on X Thursday. “The EU should be supporting free speech not attacking American companies over garbage.”  “Much appreciated,” Musk responded.  A key aspect of the alleged violation is how Musk handles account verification on his social media site. Musk’s X “allows users to subscribe to a tier of the platform that grants them a badge that had previously signified the person had been vetted and approved by X’s moderators,” the Washington Post reports. The European Commission said this system makes it “difficult for users to judge the authenticity of accounts and content they engage with.” “This deception,” the body continued, “exposes users to scams, including impersonation frauds, as well as other forms of manipulation by malicious actors.” The commission also said X didn’t provide a transparent advertising repository, as the Digital Services Act requires,  and “fell short of an obligation to let researchers access and analyze its public data,” per The Post.  It doesn’t look like Musk will face similar issues in the US.  Federal Communications Commission Chairman Brendan Carr claimed on X that, “Europe is fining a successful U.S. tech company for being a successful U.S. tech company.” Musk reposted. “The European Commission’s $140 million fine isn’t just an attack on @X, it’s an attack on all American tech platforms and the American people by foreign governments,” Secretary of State Marco Rubio wrote, adding, “The days of censoring Americans online are over.” Musk endorsed the post with a one word reply: “Absolutely.”  > Once again, Europe is fining a successful U.S. tech company for being a > successful U.S. tech company. > > Europe is taxing Americans to subsidize a continent held back by Europe’s own > suffocating regulations. pic.twitter.com/EzeOWZRC2t > > — Brendan Carr (@BrendanCarrFCC) December 5, 2025 While Musk and his supporters herald X as a bastion for free speech, his tenure in the past few years has been more complicated.  In December 2022, Musk suspended the accounts of several high-profile journalists—from outlets like CNN, The New York Times, and WaPo—after Musk claimed reporters were endangering his safety by sharing information on where his private jets were using publicly available data. “Criticizing me all day long is totally fine, but doxxing my real-time location and endangering my family is not,” Musk posted at the time. According to self-reported data, from the date of Musk’s takeover to April 13, 2023, the social media site fully or partially complied with 98.8 percent of takedown requests submitted by governments. Turkey was responsible for half of all the takedown requests, followed by Germany at 26 percent and and India at 5 percent, as reported by Al Jazeera.  During the 12-month period before Musk took over the site, Twitter fully complied with 50 percent of these kinds of requests, and partially complied with 42 percent.  Since the EU commission announced the fine, Musk has been using his X page to amplify critiques of the commission’s decision. “Total war on free speech,” one post Musk reposted read. “It’s real simple,” Peter Imanuelsen, a well-known far-right voice in Sweden, began in another, “The EU fined X €120 million because this is where the mainstream media narrative gets exposed.” Musk quoted the post with the 100 emoji. 

Late last week, the X social media platform rolled out a new “location indicator” tool, plans for which had first been announced in October. Suddenly, it became much easier to get information on where in the world the site’s users are actually posting from, theoretically helping to illuminate inauthentic behavior, including attempted foreign influence. > “It is clear that information operations and coordinated inauthentic behavior > will not cease.” As the tool started to reveal accounts’ information, the effect was like watching the Scooby Doo kids pull one disguise after another from the villain of the week. Improbably lonely and outgoing female American GI with an AI-generated profile picture? Apparently based in Vietnam. Horrified southern conservative female voters with surprising opinions about India-Pakistan relations? Based somewhere in South Asia. Scottish independence accounts? Weirdly, many appear to be based in Iran. Hilarious and alarming though it all was, it is just the latest indication of one of the site’s oldest problems.  The tool, officially unveiled on November 22 by X’s head of product Nikita Bier, is extremely simple to use: when you click the date in a user’s profile showing when they signed up for the site, you’re taken to an “About This Account” page, which provides a country for where a user is based, and a section that reads “connected via,” which can show if the account signed on via Twitter’s website or via a mobile application downloaded from a specific country’s app store. There are undoubtedly still bugs—this is Twitter, after all—with the location indicator seemingly not accounting for users who connect using VPNs. After users complaints, late on Sunday Bier promised a speedy update to bring accuracy up to, he wrote, “nearly 99.99%.” As the New York Times noted, the tool quickly illuminated how many MAGA supporting accounts are not actually based in the US, including one user called “MAGA Nation X” with nearly 400,000 followers, whose location data showed it is based in a non-EU Eastern European country. The Times found similar accounts based in Russia, Nigeria, and India.  While the novel tool certainly created a splash—and highlighted many men interacting with obviously fake accounts pretending to be lonely, attractive, extremely chipper young women—X has struggled for years with issues of coordinated inauthentic behavior. In 2018, for instance, before Musk’s takeover of the company, then-Twitter released a report on what the company called “potential information operations” on the site, meaning “foreign interference in political conversations.” The report noted how the Internet Research Agency, a Kremlin-backed troll farm, made use of the site, and uncovered “another attempted influence campaign… potentially located within Iran.”  The 2o18 report was paired with the company’s release of a 10 million tweet dataset of posts it thought were associated with coordinated influence campaigns. “It is clear that information operations and coordinated inauthentic behavior will not cease,” the company wrote. “These types of tactics have been around for far longer than Twitter has existed—they will adapt and change as the geopolitical terrain evolves worldwide and as new technologies emerge.”  “One of the major problems with social media is how easy it is to create fake personas with real influence, whether it be bots (fully automated spam) or sockpuppet accounts (where someone pretends to be something they’re not),” warns Joan Donovan, a disinformation researcher who co-directs the Critical Internet Studies Institute and co-authored the book Meme Wars. “Engagement hacking has long been a strategy of media manipulators, who make money off of operating a combination of tactics that leverage platform vulnerabilities.”  Since 2018, X and other social media companies have drastically rolled back content moderation, creating a perfect environment for this already-existing problem to thrive. Under Musk, the company stopped trying to police Covid misinformation, dissolved its Trust and Safety Council, and, along with Meta and Amazon, laid waste to teams who monitored and helped take down disinformation and hate speech. X also dismantled the company’s blue badge verification system and replaced it with a version where anyone who pays to post can get a blue checkmark, making it significantly less useful as an identifier of authenticity. X’s remaining Civic Integrity policy puts much more onus on its users, inviting them to put Community Notes on inaccurate posts about elections, ballot measures, and the like. While the revelations on X have been politically embarrassing for many accounts and the follower networks around them, Donovan says they could be a financial problem for the site. “Every social media company has known for a long-time that allowing for greater transparency on location of accounts will shift how users interact with the account and perceive the motives of the account holder,” she says. When Facebook took steps to reveal similar data in 2020, Donovan says “advertisers began to realize that they were paying premium prices for low quality engagement.” The companies “have long sought to hide flaws in their design to avoid provoking advertisers.” In that way, X’s new location tool, Donovan says, is “devastating.”

Il giornalista Brian Merchant ha ricostruito la storia dell'automazione nel mondo del lavoro, dalla prima rivoluzione industriale fino all'avvento dell'intelligenza artificiale. "Innovazioni" utilizzate per frammentare la forza lavoro e ridurne i costi. Ma resistere è possibile. A partire dalla scuola, perché "lasciare entrare l'Ai in classe è davvero un patto con il diavolo". Il dialogo con Stefano Borroni Barale, curatore della rubrica "Scatole oscure" L'articolo Il sangue nella macchina. All’origine della ribellione contro Big Tech proviene da Altreconomia.

This story was originally published by the Guardian and is reproduced here as part of the Climate Desk collaboration. Millions of Americans face having to ration heating this winter as the US federal government shutdown and mass layoffs by the Trump administration cause unprecedented delays in getting energy assistance aid to low-income households, a group that helps people pay energy bills has warned. Congress approved about $4 billion for the Low-Income Home Energy Assistance Program (LIHEAP), after Trump’s proposal to cancel the life-saving heating and cooling scheme in this year’s budget was ultimately unsuccessful. But with winter fast approaching, lawmakers have failed to reach a funding deal and appropriations remain stalled, which threaten to leave the most vulnerable families without critical energy aid as electricity and gas bills surge. “No family should be forced to choose between heat and food because of a federal funding delay,” said Mark Wolfe, executive director of the National Energy Assistance Directors Association (NEADA), which represents the state directors of LIHEAP. “If the money isn’t released soon, it will cause real harm and people will suffer.” LIHEAP is a chronically underfunded bipartisan program that helped almost 6 million households keep on top of energy bills last year, reaching only 17 percent of those eligible for assistance even before the current chaos. Due to the seasonal nature of the program, previous administrations have typically allowed 90 percent of the LIHEAP funds to be distributed by the end of October—even while lawmakers wrangled over the annual appropriations bill. This is year is different thanks to Trump’s “department of government efficiency” (DOGE). Even if the continuing resolution—or short-term spending fix—were to be agreed this week, states and tribes would probably not receive the funds until early December at the earliest due to unprecedented staff shortages. Earlier this year, the entire staff running the decades-old bipartisan program was fired—as part of the Trump administration’s so-called “efficiency” drive which was overseen by the billionaire Republican donor Elon Musk. This left no technical staff to apply the funding formula, which determines how much each state and tribe receives, and approve states’ plans on how the money will be allocated to households. The Guardian understands that the Department of Health and Human Services (HHS) led by Robert F Kennedy Jr. had been using external paid consultants and staff from other programs, some of whom were fired earlier this month. With no indication that the government shutdown will end any time soon, the NEADA is urging utilities to immediately suspend disconnections for overdue bills—until the federal chaos is resolved and LIHEAP funds are released. “Utilities must act in the public interest and pause shutoffs until federal aid is available again,” said Wolfe. In the first eight months of this year, New York’s monopoly energy provider alone disconnected 111,000 households. The national total is expected to hit 4 million shutoffs in 2025—up from 3 million in 2023, according to analysis of utility-reported data. Trump declared a national energy emergency on his first day back in office, pledging to ramp up fossil-fuel production and slash regulations to bring consumer energy bills down. In the past year, electricity bills have risen more than 15 percent in 10 states plus the District of Columbia, with the highest jumps in Illinois (28 percent), Indiana (25 percent) and JD Vance’s home state of Ohio (23 percent). The price hike is mostly down to the rising cost of fossil gas, utilities passing on the cost of investment in transmission and distribution systems to consumers, and the rapid unchecked growth of datacenters, which is increasing demand for electricity. According to NEADA research, the cost of home-heating this winter is expected to rise by an average of 7.6 percent, increasing from $907 last winter to an estimated $976 this year. About 21 million households—one in six—are currently behind on their energy bills. Household energy arrears rose by more than 30 percent, from $17.5 billion in December 2023 to $23 billion by June 2025. A health department spokesperson said in a statement: “The Democrat-led shutdown is preventing states from receiving new funds under the Low-Income Home Energy Assistance Program (LIHEAP). The Trump Administration is committed to reopening the government for the American people.”