請負業者のAIエージェントが国防総省に侵入した経緯、Anthropicの終末論的IPO、投票所のICE

AI can make businesses faster. It can also make mistakes, expose data, and execute the wrong instructions faster. ThreatLocker helps organizations set boundaries AI cannot ignore by controlling what applications can run, what they can reach, and how they interact with sensitive resources. Innovation needs more than good intentions. It needs limits. Control AI with zero trust at threatlocker.com slash wired. That's T-H-R-E-A-T-L-O-C-K-E-R dot com slash wired. Welcome to Wired's Uncanny Valley. I'm Brian Barrett, Executive Editor. I'm Zoe Schiffer, Contributing Editor. And I'm Leah Feiger, Director of Politics and Science. This week, Amdropic has officially filed to go public. There is a lot going on in the hundreds of pages that the company submitted as part of its S1 filing, from staggering earnings and losses to warnings of AI's existential threat.

Obviously, we're talking about Anthropocare. We'll discuss our key takeaways and why this IPO is kind of high stakes, not just for the AI industry, but honestly the entire U.S. economy. We're also bringing you a scoop about how a defense contractor plugged Anthropocode into a secure Department of Defense system in what's really one of the most outlandish uses of AI in the government we've seen so far. And the midterms are right around the corner, just a few days away. Massive groups of volunteers all over the country are already getting ready to protect their communities from the potential that ICE officers show up at the polls. We'll get into how grassroots efforts like these are a reflection of a deeply uncertain election cycle. So on Monday, we learned that Anthropic finally pulled the trigger and publicly filed for its IPO. This has been highly anticipated for a bunch of reasons. Anthropic is expected to raise more money in its IPO than any company in history. And it would be the first major U.S. company focused solely on building AI software to go public.

To give you a sense of the scale that we're talking about, the company's valuation went from $183 billion a little over a year ago to $965 billion in its most recent round. And some expect that it could be worth close to $2 trillion once it starts trading. So the filing basically gives us the first real look inside Amtropik's business because it's a private company right now. It doesn't have to report a bunch of stuff. And the numbers are pretty extreme in both directions. Billions of dollars in revenue and even more billions in losses and hundreds of billions that the company has already committed to compute. Because there's a ton to unpack in this paperwork, I thought it would be fun to kind of divide it into green flags, yellow flags, and red flags that we saw from the filing. What do you guys think? Yes. I'm astounded by how much money is at play here. This is, yes. What flag, Leah? What flag is that? I don't know.

For the sake of the economy, it's green. Great. Green flag. For the sake of the economy, it's green. This is so much money, guys. I mean, sue me, but the only thing I can think of right now is that article that Wired ran a couple months ago about how people were selling their homes in the Bay Area with the promise of Anthropic stock post-IPO. And I remember at the time I was like, how close are we to this? And Zoe said, oh, they're going to make so much money. Is it fair to say, Zoe, I feel like having a financial effort, yes, they're losing money hand over fist. But the financials, I think, are better than people were expecting, maybe. At least it feels like they put a lot of concerns to rest, not all of them. But this is a real business. This is a business that is growing. It's doing all the things you would want a business to do. Yes, I think absolutely. So its revenue hit $16.3 billion in the first half of 2026. And it's grown more than 14-fold from the same period last year.

I think we have to go back to 2022 and remember that when Chat2DT came on the scene, it wasn't even apparent that like AI would have a real enterprise application. Like this was a consumer app that didn't have really any business case for it. And it wasn't until Anthropic came on the scene years later, it rolled out Claude Code that we started to really see like, oh, no, there's a viable business. And actually, the majority of the money is going to come from other businesses using their coding product. This is also, I think, a key difference that you hear from insiders between Dario and Sam. People very close to Dario Amadei, the CEO of Anthropic, will say he wakes up every single day and says enterprise code. That is the thing he cares about, and that's because that's the thing that makes money for these businesses. Whereas I think the perception from some is that OpenAI is a little more scattered and all over the place. And I think you can see that in the numbers. So they are leaning in. Yeah, they have like cloud code, huge success. Enterprise code, huge success.

I will say, to introduce a yellow flag, of that revenue, right, that you talked about, like very impressive revenue over $16 billion, nearly half of it came from three tech companies' cloud marketplaces. So they are pretty heavily reliant on a small number of companies, which are also competitors, right? It is this sort of tangled nest of reliances in this industry that feels, I think it wasn't, Anthropoc didn't put it as like one of its major risk factors. We can talk about the risk factors in a second because that was my favorite section. But it does feel like, you know, where's the moat? Like, is there enough of one given how much they're relying on a handful of companies that are also developing their own AI? Other AI products are getting better. Open source products, open weight products are out there. How yellow is that flag? I think it's yellow. It's not surprising that a polyamorous worker cooperative would be incestuous in its business dealings.

I was not sure. And it's not unusual for a company, for like a small percentage of spenders to have an outsized impact on a company's revenue. I think there was another number that Ramp put out recently. 80% of revenue for both Anthropic and OpenAI comes from about the top 1% of AI spenders. So the companies that are spending the most, you know, account for the vast majority of the company's revenues. But it is a vulnerability because if there's a pullback and those companies start spending less, Anthropic is exposed. But another, I don't know where to put this back. I think it's green. They were transparent about some of the risks involved in the IPO, some of the kind of existential risks of AI generally. This was probably my favorite part in the whole thing. Quote, concerns have been raised, dot, dot, dot, that sufficiently advanced AI systems could, if not developed and safeguarded, appropriately pose catastrophic or existential risks to humanity.

Incredible stuff. The use of the passive voice in this S1. Concerns have been raised by Dario. Yeah, by us. Like, internally. By anthropocent. So I will say, too, the phrase existential risk comes up three times. It's twice existential risk to humanity, once existential risk to society. So in that scenario, presumably, there are still humans. They have just abandoned any semblance of a social contract. Great. Zoe, I love that this was a green flag for you, the fact that they're so openly. She's a transparent lady. To be clear, they've been shouting this since they started, that AI could kill everybody. This is not new for them. And what I thought was interesting, and I'll call it a red flag, red flag for humanity, yellow flag for anthropic, green flag for me because I love to talk about it. So the same thing that they cite as most likely to cause the end of human life on Earth is the thing that they also cite as one of their primary competitive advantages.

And that is recursive self-improvement. So that's where an AI model can improve itself. And they're saying on the one hand, if AI models keep getting better at improving themselves and can eventually improve themselves to the point where humans don even need to look at it closing the loop where AI just gets better and makes AI better and makes AI better To the point where we don even understand what AI is doing anymore But gosh it good and it making a lot of money That's the scenario in which AI is powerful enough to maybe end the world. But it's also, Anthropoc, you know, it's also really good at that. We're investing a lot of money in that. That's like how we're going to get ahead. And how relatable is that? Who among us doesn't have a shadow side that is equally augmented by the power that that shadow side gives you? To be clear, and Wired covered this, as did basically everyone, when former anthropic researcher Jacob Coxman resigned and said it was because anthropic could bring about the end of the world. He specifically cited curbing RSI at the top of his list, right? Like this was, this is the whole ballgame.

Yes. And then the industry in general is very, they're even, I think this is probably too extreme, but like Bernie Sanders has a bill out there and Ro Khanna has a separate bill that say, hey, let's just ban recursive self-improvement. Let's just like say we're not doing that. I will say one little note is that the risk factors is the longest section of the prospectus. It's 94 pages, I believe. I guess they could have just said, we have a shadow side and saved everybody. I would have understood that. Can we talk about a red flag that I haven't gotten a great answer on yet, and I kind of need you to explain this to me like I'm a toddler? The compute bill. All of this technology costs so much money. I don't personally feel like I have an understanding of where the limit is, how you do not spend enough to use all of your money and also spend enough to not put yourself out of business and end your technology forever.

Perhaps I do not understand compute in its entirety or where it comes from. But like all of this together, this is just so much. Tell me. It is so much. I'm not going to pretend that I understand it. Like the scale of what we're talking about is, again, totally unprecedented. You get a lot of railroad analogies when you start talking about this stuff. But I will say that one key difference is that unlike the railroad companies, which were operating on slightly thinner margins, Like these are still very well-funded and rich companies. And so, yes, they're investing a ton in compute and they expect that that compute will be needed. So far, it is. And they expect that to continue to rise. I personally think that that is a viable argument. But of course, the flip side is like, well, if growth starts to seriously slow down and they're left holding the bill for all of this compute that they've committed to, a lot of the contracts that Anthropica scientists can't get out of, that's going to be really bad.

And it's going to be bad not just for Anthropica, but for the entire U.S. economy. I mean, they call it their most significant operating cost, in quotes. I mean, yes, absolutely. I think significant, too, that unlike, say, SpaceX, unlike Meta, other companies that are heavily invested in AI, a lot of what Anthropik is doing is basically leasing compute. Like, they are not building as much. They are contracting out. So, you know, on the one hand, that maybe frees them from long-term, you know, obligations on these massive properties. On the other hand, obviously, it's sort of, to Zoe's point, there are contracts they can't get out of. They are committed to these contracts, and they don't really control their own destiny in that regard, especially when the compute is very competitive now. Getting access to compute has been really, really hard. It's getting harder, especially as finance companies are less inclined to provide the debt, to support the debt, to pay for all these infrastructure costs. Yeah. One thing I wanted to add really quick is that they do say in the prospectus, I just need to call this out,

that the company needs to make forward investments that will seem, quote, irresponsibly large at first, which is Zoe found the shadow self-relatable. I found this relatable because it is what I tell my family every time I order DoorDash. I know. It's going to be irresponsibly large at first glance. But we'll pay off eventually. Yeah. Absolutely. I want to move us over to talk about a massive story that Wired published today that also has to do with Anthropics Claude. Wired's Vittoria Elliott reported today all about a security incident at the Department of Defense caused by a contractor who connected a secure centralized system, one used by multiple intelligence agencies, to their personal account for Claude Code, Anthropics commercial AI coding agent. The contractor works for the global consulting firm Accenture, and according to sources, the agent sent information about the DoD system, along with login credentials, to a high-risk third-party website.

And it ran inside the DoD network, sources tell us, for two months before anyone caught on. No. Red flag. Red flag. Red flag. Red flag. I mean, really, to start us off, and I want to get into these details, and everyone should go read this on Wired.com, But I think the overarching thing that I really want to get across here is we have not seen this before. This is a massive, massive example of an AI agent, in theory, wreaking havoc or even just having access to government systems without people's knowledge of that. First of all, huge scoop by Tori. Major credit to her. Really, really well done. I'm surprised there was not a system in place to catch this. I'm surprised that it was allowed to happen in the first place. It feels like there are multiple levels of failings here and I think speaks to just sort of, you know, this is the duty. This is happening, I assume, at all kinds of organizations as people become more reliant on agents to do literally whatever for them. Yeah. God, those people.

Leah, my question for you, and then I want to give my take, is that we don't know that anything nefarious was going on. This contractor was trying to use Claude to write code and deploy it. Yes. So this isn't like a massive hacking attempt. I think that's more interesting, though. I agree. I agree. It's the idea that there was access at all. Yeah. The thing that it reminds me of is a really popular product in Silicon Valley, maybe everywhere right now, is this recording tool and transcription tool called Granola. Basically, it's this thing that connects to your calendar. When you have a meeting, it automatically starts recording. The result of that is that in many states, You're illegally recording people all the time. And it's become such a norm in Silicon Valley that everyone's just kind of thrown up their hands and been like, whatever, like, I guess this is just what we're doing now. And this kind of reminds me of that. Like, I think that people are so used to having agents do things for them or using cloud code

where they would have previously written and deployed code by hand that they don't, they're not as aware that what they're doing is breaking norms, rules, laws. I just want to quickly confirm because I Googled it just now. California is a two-party consent state. If that's useful to anyone out there who's using granola. It's fully illegal. It's fully illegal. You are breaking the law. I really wish that you could see my slack-jawed, open face right now, the fact that that exists. But back to the Accenture of it all, back to the DOD of it all, this system that this CLAWD agent had access to, the centralized system contains controlled unclassified information. So it's CUI, it's considered to be sensitive, but not to be released to the public without review. The issue, though, is that the way that different pieces of this information from this system and other parts that the system has access to can be combined to create classified information.

Does that make sense? Yeah. Unfortunately, yeah. Okay. To spell this out a little bit more, this can include information about critical infrastructure across the U.S., nuclear submarine propulsion, vulnerabilities to chemical and biological terrorism. Like, this is not casual stuff that they had access to. The source aware of the incident who requested anonymity to speak to Tori said to her it puts all the data that in there at risk If the third software was compromised there could be somebody out there with the ability to log into a secure environment without anyone knowing It's 100% a national security threat and could cause irreparable harm. This is serious. So the Accenture employee who was alleged to have connected the secure production environment to their personal clot account to allow this agent to write and deploy code is also breaking another rule technically because Claude is currently barred from being used by DOD staff. Oh, yeah. Were we ever so young? Ever so young. This is a Mad Libs story.

This is a Mad Libs story. To automate the process, the contractor gave their login credentials. The agent wasn't just writing code but committing it, meaning that it was like saving the code to the system on the employee's behalf. And also, they had to connect to the Internet. So some of these things were air-gapped. Do we know what kind of accountability looks like here? Like what has happened to the employee? What Accenture or the DOD are doing in the wake of this? That is a great question. Despite the security risks, sources told Wired that Accenture employees did not really treat the incident as super severe or urgent. Good, good, good. Yep, good, good, good. So when we reached out to the Department of Defense, a spokesperson there told us that, quote, the department does not comment on the status of its networks or systems. And on Accenture's end, their spokesperson said that they, quote, take all concerns seriously. These allegations were thoroughly and independently investigated and found to be without merit.

We fully complied with our legal and contractual obligations. When pressed, though, Accenture didn't offer clarification on who conducted the investigation or dispute that there had been what could be characterized as an incident. And this is, I have to say, this isn't the first time that Accenture has had a bit of an issue recently. Earlier this month, the FBI found that an Accenture contractor did not patch a vulnerability that led to the hacking group Shiny Hunters getting access to all of the FBI data. I don't know if you guys remember that fun story. So lots of things happening here. So many different questions, for me at least, brought up about, like, the use of personal AI agents when it comes to the workspace, contractors in government and what they have access to, what they have to reveal to the government at what time. There's a lot here that fell in the cracks. And like I said earlier, Mad Libs, Mad Libs of a bad situation just really gone awry. I feel like the use of AI agents in general is a shortcut, right?

It's like I want to take a shortcut to whatever it is I'm doing. Sometimes that's for great reasons. Zoe, you've given over 80% of your life to agents, and you're thriving. But there are some situations where you really shouldn't take any kind of shortcut, especially not a shortcut that you can't really control that's going to be touched in sensitive documents. It's inexcusable. I assume not the last time we're going to hear about something like this. AI tools can write code, access files, connect to the internet, and execute commands at a speed no human can match. That productivity is powerful, but it also means one bad instruction or compromised tool can quickly become a security incident. ThreatLocker helps organizations use AI without losing control. Put zero-trust guardrails around tools like clawed code and control exactly what they can execute, access, and connect to. Set clear boundaries around the applications, files, systems, and online destinations your AI tools are permitted to use.

Because a proven AI tool should not mean giving it unrestricted access to your environment. Control what runs. Control what it can do. Control what it connects to. Learn how ThreatLocker can help you safely embrace AI at threatlocker.com slash wired. That's T-H-R-E-A-T-L-O-C-K-E-R dot com slash wired. Well, in other news, the midterms are officially a week away and early voting already underway in lots of states. You may have already voted as you're listening to this power to you, if so. One of the questions hanging over these elections, though, has nothing to do with who's on the ballot. It's who's going to be at the polling stations. For the last six months, President Trump and his allies have declined to rule out the possible deployment of ICE agents to voting sites during the midterms, even though that would be illegal under federal law has not really stopped him before in various scenarios. This week, Wired's David Gilbert spoke to a dozen people or so taking part in a nationwide effort to get ready for that possibility. They're in towns and cities across the country, and they are people who see the presence of ICE at the polls as an existential threat to democracy.

They've built massive networks of people to do a couple different things. The overall goal, though, is to, A, document what ICE is doing, and B, spread the word about rights to voters. So if you do go to the polls and you see ICE there, you know how to respond and react. Great that people are planning for this. Leah, I'm curious, how much of a likelihood or inevitability do you think this is going to be, ice showing up at the polls and kind of disrupting, I guess, their very presence is disruptive, right? I think so. It's really just how likely do you think it is that they actually show up? A really good question and one that I'm honestly changing my mind on every few days. Recently, the Trump administration has indicated that they will not be sending ice to the polls. Like you said, the last couple months have been the exact opposite of that. And then you also have DHS and other random government accounts posting about how they are apprehending illegal voters as we speak. So there's some mixed messaging that's clearly happening right now.

And in the middle of that, these organizations have stepped in to say, okay, here's our plan should this happen. Simultaneously, I think there's just a lot of fear. Do I think that every single polling station is going to have that? No. Do I think that fear about it is probably going to dissuade some people from voting? Yeah. Yeah, I do. This isn't forcing you to stand in a line for hours on end, which, to be clear, also bad. This is the threat that you are going to be thrown in a facility and never see your family or your home ever again. That's like this is this is massive. This is this is what ICE tactics have been in the past. So applying it to this situation, very, very scary. Will it happen? I think we're probably going to see some of it. I think it's going to happen in like a couple of different ways, whether that's agents standing near polling places that act as an intimidation to certain voters. Are they interfering or not? Tough to say. I think that there's also going to be a lot of probably false accusations of this as well. We don't know. But the possibility for misinformation, I say misinformation, not disinformation, because people are just going to be spreading things around that they haven't perhaps verified because of fear, because of what the Trump administration has been promising and has done in so many communities across the U.S. over this past year, is very real and very scary.

Yeah, I think that's exactly right. I mean, I think fear is the point and confusion is the point. I live in California. We already have multiple examples of people we know who probably won't be able to participate in the way that they normally would have because the possibility that ICE would be there is just too scary. and so whether or not they're actually at a particular polling place it might not matter and I think if they are at a few I think we can expect to see a kind of flood the zone where those images start to go everywhere and in some cases it's people trying to warn other people but I would expect that also in a lot of cases it might be just you know people trying to spread fear and dissuade people from getting out to vote. And that's where some of these organizations come in that David talked to, right? You have 614 Ice Watch. You have groups like the Common Cause, ACLU, Indivisible, that have been training their teams and these massive groups of volunteers to actually be able to track examples of ICE at the polls

and to share verified information with their networks, which include, you know, people that have been working in democratic spaces for a very, very long time. So that is in place. It's going to be about wading through the noise. I think 614 Ice Watch is a really good example because they focusing their efforts in Columbus Ohio home to a large Somali community that was targeted by an ICE operation last year This is a community that's really rattled by ICE, is really affected by ICE. And so even just knowing, look, there's someone here who's looking out for you. There's a group of people here who, if something does happen, they'll be there. Whether ICE shows up or not, it may encourage people to go to the Bulls when they otherwise might have stayed home for the reasons Zoe was just talking about. So even just it's sort of a preemptive strike that I think is really great. It's great that they're doing it. And I hope that it is effective. I hope people still are able to go out and vote. Absolutely. And to Zoe's point, because it's not me if I don't bring this into some of the political machinations behind the scenes as well.

I think that this is in the fear around it and the discourse around it and all of this is just is really just heightened by the fact that Republicans are getting trounced in the polls. you guys, like absolutely trounced. Wired senior correspondent Hugo Lowell put out a newsletter this week all about how Republicans the week before midterms are doubling down their spending and resources on Senate races in states that are traditionally deep red. We're talking Kansas. I'm sorry, in what world is that up for grabs? And it's not even to say that that election is like these elections are necessarily for sure absolutely going to go Democrat, it's that they're frightened. It's that the polls are way closer than they should be. And so all of this aside, what's a great way to start for the Republican Party to start pointing fingers, oh, this election has been impacted by voting of, quote, unquote, illegal immigrants. Oh, we're going to send ICE. ICE should be there, X, Y, and Z. So it's all of this together has created an environment of real uncertainty.

It's a very apt time to have the elections days after Halloween. I'm spooky, spooky all around. Wait, Leah, because it's our last recording before the midterms, what is your prediction? How are you going? House, Senate, who's taking it? Dems are going to take the House. Senate is still a toss-up to me. Red flag, red flag. You've got to pick, Leah. Hey, that's pretty good. I feel like I gave a good guess there. All right, all right. Coming up after the break, we'll share our Wired Tired Picks for the week. Stay with us. AI tools can write code, access files, connect to the Internet, and execute commands at a speed no human can match. That productivity is powerful, but it also means one bad instruction or compromised tool can quickly become a security incident. ThreatLocker helps organizations use AI without losing control.

Put zero-trust guardrails around tools like Claude Code and control exactly what they can execute, access, and connect to. Set clear boundaries around the applications, files, systems, and online destinations your AI tools are permitted to use. Because a proven AI tool should not mean giving it unrestricted access to your environment. Control what runs. Control what it can do. Control what it connects to. Learn how ThreatLocker can help you safely embrace AI at ThreatLocker.com slash Wired. That's T-H-R-E-A-T-L-O-C-K-E-R dot com slash Wired. Okay, it's time for our Wired Tired segment. As usual, whatever is new and cool is Wired, and whatever passe thing we're over is tired. Let's see what we've got. Zoe, I've heard you got a good one this week. Yeah, I dedicate it to my dear friend Leah Fiker. As you are well aware, as everyone is well aware,

the billboards in San Francisco talking about AI, I think, are insane. They are not legible as any part of the English language. Unclear who they're talking to or why they're talking at all. I'm sick of them personally. It's like complete nonsense and confusing. Yes, it's gobbledygook. driving through San Francisco is hell. Yes, it is. In contrast, The Row took out a double page ad, an FT. I know. I was like, you know what I'm talking about. A few weeks ago, and it was just a William Eggleston photo, just like a landscape photo, literally no like text or anything. And I was like, way to zig while everyone else is zagging. It just spoke to their taste and aesthetic. I thought it was so fucking classy. I loved it. I'm so with you talking of ads in unexpected places. Did you see that coterie? I don't know if I'm pronouncing that correctly. The diaper company. They took out an ad in Gulf Digest as a way to target the uneven mental load in parenting.

I loved it. I do like that. And it's also a great ad. Yeah. Wow. Some good ads. All right. That's a good. That's a really good wire. Big fan. Any tires for you? The hideous billboards. That's what you discussed. That makes sense. Absolutely, yes. Brian, how about you? Sure. I'll be quick. I think my wired, not because of the thing itself, but the only person gets squeezed in some consumer tech news in this podcast. Apple is releasing their first touchscreen Mac. They said they'd never do it. They finally did it. I personally don't care about a touchscreen Mac. I will never buy one. I don't have much use for one, but I like that they're at least doing something a little bit different, and I like that they have gone back on their word. That's fun. That's wired. Am I tired is going to be, you all may remember, when Chris Nolan's The Odyssey came out. Big deal, great movie, a lot of fun. Elon Musk and associates, not fans because it was too woke

in various ways. He promised that Grok would make an un-woke Odyssey and by God, they did it. Thank God. They released the first chapter of chapter, whatever, seven minute snippet of the Odyssey. And boy, is it terrible. It's it's it's extremely bad. I don't think it's woke, at least. I think they succeeded on that front, but not not super watchable either. So that's tired for me. That's pretty good. Leah, what about you? So my tired this week is, this is so boring of me, but election disinformation about quote-unquote illegal voters. This is just such a, this is a thing that I'm going to talk about for the rest of my life. The number of people voting who are not allowed to do so in the U.S. is such a teeny, teeny, tiny number so as to have certainly never once influenced an election. And I am so tired that day in, day out, election in, election out, I am still talking about this.

It's tiring. I'm over it. Go read all of our reporting on Wired.com about how wrong it is. My Wired for this week is entirely unconnected. Halloween is on a Saturday. Halloween is not going to be on a Saturday again for many years. Really fun. I have some good plans. I hope you do, too. I can't wait to see all your costumes. Brian, Zoe, send your children's costumes. Oh, yeah. I was going to say, I'm an adult. I don't wear costumes, but my kids. Absolutely. We are still nailing those down, but as soon as we get it. DoorDash, Brian. I will send it right over. Perfect. That's our show for today. We'll link to all the stories we spoke about in the show notes. Adriana Tapia produced this episode. It was mixed by Cran Vandy, who's also our New York studio engineer. It was fact-checked by Matt Giles and Daniel Roman. Kate Osborne is our executive producer, and Katie Drummond is our global editorial director.

AI can make businesses faster. It can also make mistakes, expose data, and execute the wrong instructions faster. ThreatLocker helps organizations set boundaries AI cannot ignore by controlling what applications can run, what they can reach, and how they interact with sensitive resources. Innovation needs more than good intentions. It needs limits. Control AI with zero trust at threatlocker.com slash wired. That's T-H-R-E-A-T-L-O-C-K-E-R dot com slash wired.