Support for this podcast comes from Is Business Broken?, a podcast from the Marotra Institute at BU Questrom School of Business. A recent episode, recorded at the WBUR Festival, asks what transforms an idea into a technology, a company, a job, or even an entire economy. Stick around until the end of this podcast to preview the episode. WBUR Podcasts, Boston. It was like something out of a spy thriller. An intrepid reporter got his hands on a government file. You might be picturing vanilla, manila folder, classified, splashed across in a red stamp, all caps. And you probably wouldn't be far off. The file did look particularly Hollywood-esque, at least to the reporter who saw it. It has like the person's name and residency and occupation. It's just it's done up in a way that makes it seem like this person is a criminal and like they are gathering, you know, tons of intelligence on them. That's Del Cameron. He's an investigative reporter with Wired covering privacy and national security. And he's telling us about the time he got his eyes on a document out of Washington state. The document was created ahead of a No Kings rally there last year. This included something that they called, quote unquote, a threat escalator section. And it was built entirely around a single person who was a Harvard grad student living in Seattle. The document is sourced entirely to a TikTok video that she posted in the days before the event. And she's focused on kind of warning her viewers, you know, her audience about impending ICE raids in the Yakima Valley. And she's saying ICE, these ICE agents are domestic terrorists and should be dealt with like domestic terrorists. So that intrigued the police and they circulated a document to at least, you know, dozens of officers, probably hundreds in the area. Dell got this particular file from Property of the People, a nonprofit dedicated to government transparency. It's what's known in law enforcement circles as a threat bulletin. And it had a lot of personal information about this Harvard graduate student. That page that they circulated, you know, included her name, her social media handle, where they believed she lived, you know, occupation. They lifted photographs of her from social media and from the video. And then interestingly, it kind of it just characterizes her speech as spreading extremist ideology, attempting a call to action against a government entity and invoking fear and panic. And then it also, you know, it ultimately concedes that this person is not even likely to attend the protest itself. And there was something else noteworthy about the file. The word threat is misspelled on this document. It looks like someone attempted to put this thing together to make it look really professional, but in the end failed to some degree. Dell told us it's not unusual for the quality of these threat bulletins to be quite poor. The information in them, mined almost exclusively from social media, can also be flimsy. Still, they're circulated to officials, and the consequences of ending up in a document like this can be pretty severe. It can put you on the federal government's radar, or it can put you on the radar of your local police department. That is something that could potentially affect you if they add a little flag next to your name in some database or something. Maybe you get pulled over for a traffic stop and you may not realize what's going on, but the officer is not treating you very well. So when Dell saw this dossier-like document about a private citizen built on a single social media post, it sent a tingle up his spine. That was probably the most insane instance of something like that I'd seen. Del has seen some pretty insane things on his feet. As a reporter, he spends a lot of time parsing out the ways the government is watching us and gobbling up our social media data. Most of us know that nothing we say online is really private. Still, the internet is a big place. There are millions of posts every day across TikTok, X, Facebook, Blue Sky, threads, etc., etc. It's easy to feel like no one's really watching you specifically. Unless you have a habit of going viral, of course. Most of us do not, right? Right? We're just yelling into the social media void? Well, Dell has some bad news for us. I don't want to tell anyone not to be on social media, but if you're using social media, you should assume that the government can read that information, at least. It gets worse. It's not just that some government agent is scrolling through your feed. It's that all that data, your profile, mine, everyone's, has been scooped up and put on the marketplace as a product the government can buy in bulk and use for its own intelligence purposes. The truth is that, like, the way that things are being architected now is that the social media companies or data brokers are really collecting information on everyone. And then the government can come along and run queries on that information. The Trump administration is very interested in running queries on our online data. Here's just a sample of some of the contracts Dell told us about. Palantir sits kind of at the center of all of this. There's a company called Zingle Labs. They supposedly process like 8 billion social media posts a day. Anytime you've posted a picture of your face online, there's companies like Clearview AI that are scraping those. They just got a $9.2 million ICE contract in September. And boy, is that just the tip of the surveillance iceberg. I'm Ben. Stop surveilling my data, Brock Johnson. I'm Anne-Marie. Social media is a slippery slope, Severson. And you're listening to Endless Threat. Coming to you from WBUR, Boston's NPR. Today's episode, socials for sale. All right, Anne-Marie, we know the government, more specifically the Department of Homeland Security and ICE, is watching and lurking on our social media data. And the question is, what for? I mean, I think the mission is to deport as many brown people out of the country as possible. You have, like, ramped up quotas over at DHS. They're trying to, like, kick 3,000 people out of the country a day or arrest them and put them in these camps. So it's used towards that. I mean, it's also used to silence dissent. I mean, they're using it against protesters. That's very clear. And they are using it as well for its chilling effect. Is the purpose to build this massively comprehensive database that just has you put in a name and it can tell you absolutely everything about a person? Or are they looking for certain key words, key sentiments, key threats? You can use it both ways. Like you can use you can use a keyword search maybe to pull a list of people out of this database. Or you can have like an individual that you want to investigate and put their name into it and pull out a ton of information about them. Put simply, the government can use our social media data to create lists of people to target or it can build dossiers on people they're already targeting. When it comes to identifying targets one of the key tools DHS is using is facial recognition If you are making comments online about the government that they don like which can be very benign they may try to identify you They have this face print. They know from that face print, like which protest you were at, maybe like what time, location, things like that. They can put a name to a face and then they can move you into a derogatory, what they call like a derogatory list, which is just a watch list of people who they have no necessarily don't necessarily have any evidence that you've committed any crime. But, you know, it's a secret watch list of the kind that the FBI has also run in the past. So what kind of benign comment online might get you flagged? You know, just just asking for a friend. The example that stands out most right now is anyone who is monitoring ICE activities like ISIS here. This is where they are. There was a lot of websites that people were circulating that were meant to like track ice activity. There were apps, websites that had maps on them. This is where ice activity was happening. That kind of stuff drove them absolutely crazy. And, you know, they were not willing to interpret it as citizens, you know, observing law enforcement activity legally. So the government can watch you, but they don't like when you're watching them. But let's go back to that facial recognition piece that Dell was talking about. Because according to his reporting, ICE agents out in the field will take pictures of people. And then they can see in real time if someone's face brings up a match in a database. But there's a big problem with that. This isn't like running someone's fingerprints or driver's license through a database to see if they have a warrant or something. Dell says the facial recognition technology being used isn't entirely accurate. And the government knows it's not accurate. But it's getting used anyway. There's a high possibility that when they take a picture of someone, they're not going to get the right result. And what we've seen in litigation and court filings and testimony over and over is that it doesn't really matter. Like, they run the face once, they'll get no match, they'll try to run it again, they may get no match, and then they just arrest the person anyway. DHS told NPR that the app they're using for these facial scans does not rely on open source media or social media. But that doesn't mean they're not using that data elsewhere. This February, Dell reported on a contract between U.S. Customs and Border Protection and a company called Clearview AI that would give officials access to billions of images scraped from the Internet, including faces for the tidy sum of $225,000 a year. Which doesn't sound like that much money, actually. Yeah, I mean, I basically pay that for my television subscriptions at this point. So we know the government is interested in facial recognition as part of its anti-immigration agenda, even when that tech is less than perfect. But there's more. They're interested in our location data, too. If the government can figure out where people are, it can use that information to better target them in immigration raids. And they're going to use third-party contract analysts to do that work for them. Last year, Dell reported on ICE's plans to station nearly 30 private analysts at agency facilities in Vermont and Southern California. Those analysts would be tasked with one job, to scour our social media feeds and drum up new targets for immigration officials, relying heavily on artificial intelligence and commercial databases to get the job done. The analysts are digging for posts where people might mention their location. Like, hey, I'm at a party at this person's house or just grabbing a sweet treat from this coffee shop I love. You're looking through your posts, seeing what neighborhoods you're in. They build social maps. So you'll have a post from someone that was at a barbecue with someone else. And, you know, they try to correlate that information to figure out, you know, where this person they're hunting is likely to be. who their friends are, who their family is, that way that they can narrow down the scope of like what physical geographical location this person is likely to be in. Once the analysts have enough information about someone's location, Dell says they'll put a targeting package together and send it to an enforcement branch who can then carry out the raid. Even if you're chronically offline and you and your friends don't generally post about your location, that location data still isn't safe from the government because it's for sale. Lots of apps on your phone collect location data, like the weather app. That data is usually used to advertise something to you, but the government can buy it too. And, well, it is. Basically, they're buying your location data from a data broker who is getting that location data from potentially your cell phone company is selling it. Or apps that you're using on your phone are collecting your location data and selling it to these data brokers. And then ICE is buying it. So they're buying a ton of that every day. Technically, your location data shouldn't be easily available to the government. Thanks to a Supreme Court case from 2018, Carpenter v. United States. The Supreme Court ruled basically that location data was sensitive enough that it required a warrant. The way that the government gets around that now is by instead of demanding it with a subpoena, they just pay the company to give them that data. If you can't do it yourself, buy it from someone who can. The government is still using subpoenas to try to get other information it's interested in. Like, for example, the email addresses and phone numbers of social media accounts criticizing ICE. That's according to reporting from The New York Times from this February. Social media companies don't always have to turn over the information, not right away at least, except... So there's a lot of cooperation with industry to hand over this, especially if it's like financially lucrative for them to do that. The other kind of aspect of it is that as social media was really becoming a thing and the Internet was kind of truly exploding through like the 2010s and into the last decade, these companies like Facebook and Google, Twitter would start getting flooded. I mean, they were just getting flooded with subpoenas and requests. And at some point, it just became advantageous for them to set up their own offices and kind of begin cooperating with law enforcement to help walk them through the process, maybe help them tell them exactly what they need to fill in to get the subpoena to go through. Something about that really bothers me. It should. It should, right? Like Twitter used to fight subpoenas like before it was owned by Elon Musk. You know, that's Twitter deciding that it wants to spend money defending its users rather than comply with the request. So it's really up to the companies now. I mean, do you think Elon Musk is going to fight those subpoenas? The thing you kind of have to think about now is that most or nearly all of the major social media companies are also defense contractors or in bed in some way with the intelligence community. We asked Dell for some examples of how social media companies are intertwined with the government, like X, now owned by XAI, which is owned by SpaceX, a major government contractor. And then there's Meta, which owns Instagram, Facebook, Threads and WhatsApp. By its own admission, Meta has made its open source AI available for U.S. government agencies, including the military and the intelligence community to use. And don forget LinkedIn owned by Microsoft another defense contractor Microsoft has a whole page on its website dedicated to how their services can support U intelligence and defense Man, it is getting a little cozy in this bed, Amory. Social media companies and the intelligence community all rolling around together. How is this? Is this a twin bed? Is it a queen? Is it a king? You know, I feel like everybody's poly, but are we actually conscious that we're poly in this bed right now? I do not know, but I think it might be time to change the sheets. You know what I'm saying? Yeah, it's getting funky. In a minute, some of the consequences of this kind of surveillance and whether there's any remaining protections on our data in this brave new world. Thank you. people continuing to believe that science matters, research matters. Follow Is Business Broken wherever you get your podcasts. And stick around until the end of this podcast to preview a recent episode in which host Kurt Nickish talks with Boston University President Dr. Melissa Gilliam about the value of fundamental research. Dean Russell here, senior producer for WBUR Podcasts. Scientist Carl Sagan once said finding truth demands the rigor of science and the freedom of imagination. WBUR's new kids podcast has both. The Midnight Rebellion uses the science of climate change to imagine a future gone wrong. And you need to set it right. All episodes are out now. The Midnight Rebellion. Listen at WBUR.org slash midnight or wherever you get your podcasts. You know, Ben, sometimes when people talk about this kind of surveillance, there's this undercurrent of like, oh, well, the government says it's going after criminals, and I'm not a criminal, and I'm a U.S. citizen. I don't have anything to hide, so it's fine. But when the slope gets a little slippery for some people, it eventually gets real slippery for all people. If you want to see a slippery slope, And last April, social media became a negative factor in the processing of immigration services. Social media became a negative factor in discretionary adjudication for anti-Semitism. So that's all tied back into the protests of the war in Gaza. And a few months later, in August, they added anti-Americanism, which is now that's an overwhelmingly negative factor for immigration services processing. And while the examples that privacy journalist Del Cameron is talking about there involve how the government is treating immigrants who are already in the country, he says the overall approach here reminds him of another turbulent time in American history. You know, that really borrows from the Cold War era bar on communists and anarchists. There's a lot of parallels between kind of the McCarthy era, the use of red squads. What they didn't have in that era, you know, was the microship. They had index cards that filled hundreds of cabinets where they would keep on people. They don't even need to build dossiers on people anymore, which is what they were doing in the 90s. And they have access to such vast databases all over that have all this information and collected in one place that they can kind of hit a button and create a dossier on you at any time. It's like more powerful and less accurate. It's less accurate. And there's no way for you to really contest what it says or possibly even know that it exists. But you can't, you know, you can't contest the inaccurate information in the dossier that the government is using to make decisions about how it's going to treat you. When we think about right now, what can't ICE do? Yeah, I mean, they can do almost everything. The things that are really protected are the contents of your messages, texts and conversations. that receives the most protection. And it receives even more protection if it's in transit, if they're trying to intercept that information live, if they're trying to wiretap you. Everything else can be bought and sold, basically. So your text messages and DMs? Probably safe. Anything else you post? Probably not. So there were some people who were in charge of making sure the government was using these tools in a way that didn't violate privacy. Small itty bitty problem, though. One of the things kind of you have to recall is that like in March 2025, that's when the reduction in force notices went out. That gutted, for instance, the Office of Civil Rights and Civil Liberties, which was the body that handled complaints about facial recognition in particular and social media monitoring abuses. There were also, you know, significant changes to the privacy offices inside of DHS. amory i think anybody who's you know reasonably aware of everything that has happened since edward snowden assumes that there's a certain amount of just broad surveillance happening of americans and people living in america whatever their citizenship status is all the time right but we're entering into this world where the government is outsourcing a ton of that work to algorithms and private companies. And the work is both bigger and broader than it's ever been. You know, we're outsourcing it outside of the government. We're outsourcing it outside of humans. But the implications feel very dystopian and also very personal. They feel like the impact of this stuff is extremely personal and potentially traumatic. And that worries me. Yeah. I mean, a lot of what Del is reporting on feels like the fine print of being an American in 2026. And a lot of people have been feeling this sort of like fine print vibe for a long time. You know, being an American is not the same from person to person. But now we're becoming more aware, at least I hope we're becoming more aware, of the ways in which we're being surveilled that we didn't even know was legal. And it feels like the rules are suddenly changing right under our feet. We're basically weaponizing data in new ways, and we're doing it with way less oversight from government watchdogs that are supposed to protect us from government overreach. But there are some people still watching. There's Del, who spends every day swimming in this stuff, which we realized is actually kind of a vulnerable thing. Yeah, do you worry for yourself at all covering this particular issue? You know, I don't want to be vain and think like the government is spying on me specifically. Like, oh, man, they must have read that thing I wrote last week. I bet they're really mad and they're listening to my phone calls. But, you know, they don't necessarily have to. I mean, the targeting doesn't have to be that specific. I think that they can, the technology, I mean, AI in particular, the technology is there now where it doesn't necessarily require a lot of people sitting around like thinking about you. You know they can make a list of journalists and run that through a system and pull out whatever information they want They not supposed to There supposed to be someone there that says no you can do that But they mostly fired all of those people. So, yeah, I mean, I definitely worry about it. And certainly in every case, if I'm dealing with like a source, like we're using the utmost protection possible. What does that look like? Is it like Signal chats and whatnot? I mean, yeah, certainly Signal is kind of the bar for encrypted communications. It's what I think most journalists use. There are other tools, but end-to-end encryption is a must. And then, I mean, but in some cases, it looks like, you know, getting on a plane and flying and sitting in a cafe and meeting someone. Yep. Which is very, you know, it sounds a little James Bond. But if you have a source and they're, you know, scared and they have a right to be scared, it can look like that. Dell says if we want things to get better, maybe they have to get worse first. You have to ask yourself whether you want your children growing up in a neighborhood where every time they go outside, they feel like they're being photographed or even recorded. I don't think that that's like the place that people want to live. I think we're trapped. I don't, you know, people use the like 1984 metaphor a lot, but it doesn't necessarily apply because to some degree we are doing this to ourselves. We're buying this technology. We're normalizing it every day. In order to participate in society, you have to, you now have to make yourself vulnerable to the state surveillance. I think the rampant abuse of it that we're seeing may accelerate the amount of pushback to it that could, you know, it could result now in some laws. So what do you think, Gamery? More laws? Should we have that? Oh, I think I think that's a piece of the puzzle of what maybe needs to happen here, because I don't have much faith that tech companies are going to regulate themselves either. No, but no, no, I don't. But I guess my silver lining is is Dell's reporting, knowing that there is someone out there talking about what's happening, even when government oversight itself is falling short. Hmm. Major headline. Journalist thinks that journalists are important. Yeah, I do. I do, too. I do, too. The point is well taken that our world is kind of like built around this stuff at this point in a really complicated way. So you can't just like necessarily, well, you can technically go and live in the woods and never, um, never engage with it. But, you know, eventually it's possible also that people will just opt out of a lot of this stuff and, and not check that. I agree box, not sign the EULA, the end user license agreement. And, um, and you know, we have that individual power to not sign it. And we have collective power and a bunch of us not signing it and not signing up for this stuff in a way that sort of changes the economy of how this stuff works and removes it as a tool that's useful for the government. Yeah. Yeah. It'd be hard for me to opt out of the weather app, but less hard for me to opt out of social media. It is possible, just saying. And who knows, maybe eventually just the awareness of the ways in which we are being surveilled could translate into some real changes of behavior and real change overall. Or at least something that makes our privacy seem less like fine print and more like something that can't be negotiated or sold. Yeah, I'm looking forward to Signal releasing its weather app. Can't wait. We're waiting. One more note before we go. We asked the Department of Homeland Security a bunch of questions, including how the government uses social media data, how agents use facial recognition technology in the field. And we put the criticism to them that this enforcement is having a chilling effect on dissent. While the agency didn't answer any of those questions directly, a spokesperson did write back to say DHS wouldn't, quote, confirm or deny law enforcement capabilities or methods, unquote, adding, quote, DHS law enforcement methods abide by the U.S. Constitution, including the Fourth Amendment. And speaking of the Fourth Amendment, the Supreme Court ruled in late June that people have a, quote, reasonable expectation of privacy when it comes to smartphone location data. The opinion came in a case about geofencing, when law enforcement asks tech companies for information about all the people whose phones were near a particular crime scene. That case was kicked back to a lower court, and you better believe we are going to be watching where that goes. So stay tuned. Endless Thread is a production of WBUR in Boston. This episode was written and produced by Kalyani Saxena. It was co-hosted by me, Amory Sievertson. And me, Ben Brock Johnson. It was edited by Dave Shaw and Meg Kramer. Mix and sound design by Marquise Neal. The rest of our team is Dean Russell, Emily Jankowski, Grace Tatter, our production manager, Paul Vykus, and managing producer, Samita Joshi. Endless Thread is a show about the blurred lines between using the weather app for forecasts in using the weather app for surveillance. If you have an untold history, an unsolved mystery, or some other wild story from the internet that you want us to tell, hit us up, endlessthread at wbur.org. And we figured we should also remind you, if you have a story that you want to share more discreetly, we are on Signal. That number is 646-456-9095. Again, 646-456-9095. You can also find that in our show notes. Thank you. And listen on for a preview of a recent episode in which Boston University President Dr. Melissa Gilliam discusses why solving society's biggest challenges requires ideas from many disciplines working together. So in universities, we run the gamut. We go from cells to society and our types of research that sets us up for entrepreneurship, innovation, company formation. It's very tempting to say, what's the return on investment? I want to see these translate quickly to society. But what you have to remember is that there's often a very long lead time to getting to those discoveries. And there's a lot of serendipity. So in universities, we focus on something called curiosity-driven research. It's a place where we allow people to pursue things that matter, even if you cannot figure out what the application is. What we're now doing at our institution is we're focusing on something called convergent research. So you use the example of photonics. You created something called neurophotonics, the combination of the study of light, but also thinking about how the study of light contributes to brain health. You created something called synthetic biology, how the insights of engineering can be used to advance biological processes. So convergence is this idea of bringing lots of fields together. Find the full episode by searching for Is Business Broken wherever you get your podcasts and learn more about the Marotra Institute for Business, Markets, and Society at ibms.bu.edu.
