Up First from NPR - Airbus A-320 Inspections, Afghan Asylum Seekers in Limbo, Real Video or AI Generated?

Episode Date: November 29, 2025

A JetBlue plane plunged uncontrollably last month prompting the parent company to issue an order to inspect all Airbus A-320 jets. It’s bad timing for holiday travel. Afghans waiting for asylum say ...they are in limbo after the Trump administration paused all asylum decisions. This comes after an Afghan national killed a National Guard soldier and wounded another. Tips on how to tell a real video from one generated by AI.Learn more about sponsor message choices: podcastchoices.com/adchoicesNPR Privacy Policy

Transcript
Discussion (0)
Starting point is 00:00:00 There's been an urgent order to inspect the Airbus A-320 family of jets. That's after a jet blue plane plunged uncontrollably last month. So what might that mean for your travels? I'm Scott Simon. And I'm Lauren Freyer. This is up first from NPR News. The Trump administration has stopped making any asylum decisions. And that's in reaction to the killing of a National Guard soldier and the wounding of another this week by an Afghan national. This puts thousands of Afghans waiting for asylum in limbo.
Starting point is 00:00:35 You'll hear from some of them. And that dancing cat video looks so real. But is it? We'll tell you how to make out a legit video from one generated by AI. So stay with us. We have the news you need to start your weekend. Support for this podcast comes from Dignity Memorial. For many families, remembering loved ones means honoring the details that made them unique.
Starting point is 00:01:09 Dignity Memorial is dedicated to professionalism and compassion in every detail of a life celebration. Find a provider near you at DignityMemorial.com. This message comes from TED Talks Daily, a podcast from TED, bringing you new ideas every day through TED Talks and conversations. Learn about the ideas shaping humanity. From connecting with your inner monologue to finding out of aliens exist. Listen to TED Talks Daily. This is Ira Glass. On this American life, we tell stories about when things change.
Starting point is 00:01:40 Like for this guy, David, whose entire life took a sharp, unexpected, and very unpleasant turn. And it did take me a while to realize it's basically because the monkey pressed the button. That's right, because the monkey pressed the button. Sprising stories every week, wherever you get your podcast. Airbus says there's a problem with the flight control systems on its A-320 family of jets. So it's ordered an immediate software switch that could ground thousands of jets worldwide for a while. Of course, that's not great timing with all the holiday travel that's coming up. NPR's transportation correspondent Joel Rose joins us now.
Starting point is 00:02:21 Joel, thanks for being with us. Hey, Scott. What does Airbus say? The company said Friday that it has discovered a problem with the flight control systems on its most popular family of jets, the A320 family. Specifically, Airbus says intense solar radiation may corrupt the data in systems that are critical to the operation of the aircraft. Airbus made this discovery after an incident last month when a jet blue plane plunged uncontrollably for a short time on a flight from Cancun, Mexico to Newark, New Jersey.
Starting point is 00:02:50 Several passengers were hurt following a sharp loss of altitude, and the flight had to make an emergency landing in Tampa. Now Airbus is notifying airlines that they need to take immediate steps to prevent something like that from happening again. Aviation regulators in Europe and the Federal Aviation Administration issued orders yesterday that airlines have to do this before these planes carry passengers again. How hard will it be to fix? Well, basically Airbus is instructing airlines to change the software on this particular computer system, either by rolling back to an earlier version or replacing the computer system with one that is running the earlier software version. It's not a difficult fix as these
Starting point is 00:03:29 things go, but it will take time, several hours per plane. In a statement, the CEO of Airbus said the fix has been causing, quote, significant logistical challenges and delays, unquote, the company apologized for the inconvenience to its customers and to passengers, but said that safety is its top priority. Thousands of planes could be affected. I'll put that into some perspective for us. Sure. So the A320 family is now the most flown plane in the world. More than 9,000 in all when you include the A319, the A320, and the A321, it's a huge part of fleets in Europe and Asia. Not quite as popular in North America, but still U.S. airlines have over 1,600 of these jets in their fleets collectively, according to the Aviation Analytics Company, Sirium.
Starting point is 00:04:14 Out of that 1600, the FAA says the emergency order applies to about 545 airbus jets in the U.S. The U.S. carriers with the most A320 family planes are American Airlines, with over 300, followed by Delta and JetBlue each with more than 200. Delta said it expects that fewer than 50 planes in its fleet will require the software fix. American said about 200 of its jets needed the fix, but that nearly all of those aircraft already had the software change completed as of this morning. Millions of people across the U.S. are expected to fly this weekend. How much will it affect them? You know, the timing is very bad for holiday travelers and for the airlines. This is one of their busiest weekends of the year, particularly Sunday with more than 51,000 flights scheduled,
Starting point is 00:04:59 according to the Federal Aviation Administration. There are 46,000 flights today, another 49,000 on Monday. So there is not a lot of extra slack in the system. Taking any number of planes out of service is going to hurt. It is just a question of how bad this is going to be. Even if it is a relatively small number of planes that are out of service, it could still result in dozens or hundreds of cancellations and delays, which then ripple across the country. as the day goes on. But that said, I think it's possible that the biggest impacts of all this will be in Europe and in Asia, where the airlines depend heavily on these planes to carry millions of passengers every day. And Pierre's Jill Rose. Thanks so much for being with us.
Starting point is 00:05:39 You're welcome. The Trump administration has stopped all asylum decisions after an Afghan national killed a National Guard soldier and wounded another on Wednesday. It's also stopped issuing visas to anyone traveling on an Afghan passport. And Pierce Dia Hadid has been speaking with some who have been waiting for years to be allowed into the U.S. She joins us now from Mumbai. Dear, thanks for being with us. You're welcome, Scott. What have these Afghans been saying to you? They've been telling me how distressed they are. These are really dark days for the sub-265,000 Afghans who advocates say were in the pipeline to come to the U.S.
Starting point is 00:06:24 Like Roshangar, he was in the Afghan military. He used to sign off on airstrikes targeting Taliban fighters. He's now in hiding in Afghanistan in fear of his life, which is why I won't use his full name. But Roshungo told me, Scott, that he was ashamed of the Afghan National behind the Wednesday attack in D.C. And he added this. That's very frustrating for us,
Starting point is 00:06:47 because that was that person's responsibility and act, death was not an act from the whole Afghan community. And that sense was also echoed by Najib Pekan. He's 45 and used to be a journalist. And he's now living and hiding in Turkey. He spoke in Dari to producer Fadi Barqbedi. And what he's saying there is, what's the difference between a man who punishes every country man for the crime of one person?
Starting point is 00:07:24 and the Taliban who imprison a family for the crimes of one's son. Dea, what can you tell us about Afghans who were in the process to come to the United States? Right. Well, most of these people worked and fought alongside U.S. forces or America's allies in Afghanistan, like Roshangar, or they were outspoken critics of the Taliban. This isn't the first time the Trump administration has tried to limit the admission of Afghan nationals. And for folks like Mishongar, that means his life has become more and more perilous. And I noticed that because as he was chatting with me, he was stuttering. And then he told me why.
Starting point is 00:08:01 I have a teeth problem. I can't go to dentists. My son is out of school for four years. We are living in uncertainty under the Taliban repressive regime. Such uncertainty. Another woman, Soraya, who was also in the pipeline to come. When I first got to know her, she was living in. Pakistan, getting ready to go to the U.S. But in June, she and her daughters were deported,
Starting point is 00:08:26 and now they live under the Taliban, too, where women are bound for most jobs, and her daughters aren't allowed to study beyond grade six. She's also in hiding, which is why I'm not using her full name. And she's one of the hundreds of thousands of Afghans that Pakistan has forcibly deported this year, including people who were meant to go to the United States. The Trump administrations of Afghans who entered the U.S. weren't properly vetted. What have you learned by talking to Afghan nationals? Well, they say they don't know what the Americans were doing behind the scenes, but someone like Rishangar said his application was in process for three years before
Starting point is 00:09:00 Trump suspended it. He'd done three interviews and was scheduled to do a fourth, and he described them to me like this. So they are asking about my background, about my family members, about my wife's family, my job, my first job, and my second job, my email addresses. my address of livings. It was so extensive, and he had to provide that information going back a decade. One prominent advocate, Sean Van Diver from Afghan Ivak, says these migrants are extensively
Starting point is 00:09:32 and exhaustively vetted. He says any gaps in the system should be fixed, but says one Afghan shooters really don't reflect the vast majority of Afghan migrants who are meant to come to the United States. Dear Hadid, thanks so much. You're welcome, Scott. Might be hard to spot a real video from one generated by AI. So we're joined now by NPR's Jeff Brumfield with some tips on how to do just that. Jeff, welcome to the show.
Starting point is 00:10:04 Hi, Lauren. So a lot of people over this holiday weekend probably have relatives and friends showing them these videos. And they're wondering, is that real? Is that AI? Just how hard is it to tell these days? It is really, really tough. these AI Slop videos. I spoke to Hanie Farid, a professor at the University of California Berkeley, who studies manipulated media, and he told me this. I do this every day, and I've been doing it for a long time, and it's really hard. It's really
Starting point is 00:10:34 hard. And to give you a sense of really how hard it is, Lauren, how about we look at a video together? Yes, I'm game. Let's go. All right. All right. So I have sent you this one. It's an altercation between a New York City police officer and two ice agents. Okay, so I'm clicking on this video. I'm looking for, like, extraneous shadows or extraneous limbs that don't fit. There's, I don't see that. Yeah, extra limbs or something, extra fingers, maybe. Yeah, but the building, the scene behind them looks normal.
Starting point is 00:11:07 I'm going to say, yeah, this is legitimate. Well, have I got news for you? This is AI generated. No. But you shouldn't feel bad because. a guy named Jan Lacoon who heads research on AI for meta and is considered one of the fathers of AI actually posted this on his social media feed, seemingly not realizing that this was AI slop. So you are in very good company. This stuff has gotten so good.
Starting point is 00:11:36 You can't look for missing limbs anymore. You know, what gives us a way, actually, according to the experts I spoke to, is think about the situation, right? Like you have a cop publicly yelling at ICE officers, and then somebody with a camera is just standing like two feet away, calmly filming the whole thing. Right. Like, how likely is that to be real? And I think that's the sort of context clues that people are starting to say you need to be looking for when you look at these videos. So should I be assuming everything I see online is fake? Actually, no.
Starting point is 00:12:12 In some ways, it's more dangerous to assume. everything's fake. And I'll tell you why. I spoke to someone named Colina Coltie. She's an investigator with an open-source research group called Bellincat. And what she does is analyze things like bystander videos of ice raids. She says if everyone starts assuming those videos are fake, society will be in real trouble. I mean, I think that's one of the bigger risk when it comes to this kind of content, right? It's not the risk that someone's going to believe a fake video, but that people won't believe real videos. And this actually has a name. It's called The Liar's Dividend because if everyone assumes everything's fake all the time, then guess who comes out on top?
Starting point is 00:12:52 It's people who are dishonest because they can just say that video of them doing the horrible thing, that was all made up. It was AI. Like it's not real. And so accountability becomes very, very difficult in a society where everyone thinks everything's fake all the time. So, Jeff, it's important that we at least try to discern what's fake and what's not. But how can I at least try to spot these AI videos? So we already talked about the context. There's some other easy stuff to look at.
Starting point is 00:13:23 One is just the length of the video. AI videos for now tend to be pretty short because it takes a lot of computing and electricity to make them. Farid says that's a really useful thing to check. When you see these little bite-sized videos, it's a good indication that maybe you should take a breath. Of course, sometimes people make longer videos, but they do that by stitching the short ones together. So usually once you get a feel for it, you can find them. Finally, all kinds of videos, it's good to look at not just the video itself, but what's going on around it. So AI videos have watermarks sometimes that you can see. Often the accounts,
Starting point is 00:13:58 you'd be surprised. They'll actually say that they post AI generated imagery. The comments can help too. People often flag in the comments that something's AI. And then similarly, real videos will appear in a different sort of feed. It'll be like a real person's account that won't just be, I don't know, cute cat videos, or it'll be a community group for a location. And then they're often picked up by news outlets. So the last thing you can do is just, you know, check the news or wait a couple hours and check the news if it's something really big. Great advice.
Starting point is 00:14:27 So, okay, you want to take one last shot out with one more video? I think I'm ready. One more. Let's see if I can do it this time. I think you can. Go ahead and check this next video I sent you. Okay. So I'm clicking.
Starting point is 00:14:38 It looks like, is that a kangaroo? Is that a moose, a deer? It's not a kangaroo. It's a donkey, and it's eating some popcorn in a movie theater. So I'm using the tools that you've just told me about. I don't see a watermark. I see a time stamp on the top. It says 908.
Starting point is 00:14:58 Reasonable time for a donkey to go into a movie theater, perhaps. Maybe the evening is showing. It doesn't look too obvious. Like, I'm not seeing the donkey. It's a moose, by the way, Lauren. It's a moose. Oh, is it a moose? But I'm not seeing the moose like up close.
Starting point is 00:15:18 You know, this looks like security cam footage. So, again, risking being duped again, I'm going to say this is legit. You are 100% right. Congratulations. You did it just exactly as you should. It does look like where security camera would be. The moose isn't in the center of the frame. It's sort of off to the side eating the popcorn.
Starting point is 00:15:40 So, yes, great work. That's NPR's Jeff Brumfield. Thanks for this info. Thanks for the tutorial, Jeff. Sure thing. And if listeners want to try this at home, we have a quiz we posted on npr.org. Take your best shot. And that's up first for Saturday, November 29th. 2025. I'm Scott Simon. And I'm Lauren Freyer. Elena Torrick produced today's podcast, along with Danny Hensel and Dave Mistich. Our editor was a D.L. Shalsh, with Jacob Fentston, Miguel Macias, Russell Lewis, Melissa Gray, and Audrey Newyon. Andy Craig, inimitably directed.
Starting point is 00:16:20 Our technical director was Nisha Highness with engineering support from Stacey Abbott and Ted Mebane. Our senior supervising editor is Shannon Rhodes. Evie Stone is our executive producer, and Luis Clemens is our deputy managing editor. Tomorrow on the Sunday story, with so many podcasts out there, we help cut through the clutter so you can pick what's actually worth listening to. For all the news you need and the stories you had no idea you needed, tune in to NPR. Find your local station at stations.npr.org.
Starting point is 00:16:53 Don't delay. Want to hear this podcast without sponsor breaks? members can listen to Up First sponsor-free through Amazon music, or you can also support NPR's vital journalism and get Upfirst plus at plus.npr.org. That's plus.npr.org. Public radio is American as Apple Pie. This giving Tuesday is our first without federal funding. We need you to keep this American tradition alive. Give now at donate.npr.org.

There aren't comments yet for this episode. Click on any sentence in the transcript to leave a comment.