No Hack Needed: How Revolut Leaked Your Passport, Plus New Age Verification & Rogue AI Agents
36 minAudio
A hacker emailed Revolut pretending to be a government agency, and Revolut handed over customers' passports, IDs, and financial details. Plus: the global age-verification wave forcing even more ID collection (EU, California, Microsoft), OpenAI's rogue AI agents and why I've hit a breaking point with these companies, and the usual run of data breaches, threats, and open-source updates. Correction: it's really DMARC that stops the spoofing, misspoke! Set up SPF, DKIM, and DMARC together. In this episode: Highlight, The hidden cost of handing over your ID: the Revolut impersonation breach, IDScan's confirmed 150M+ license leak, and the Florida DMV/ShinyHunters breach The global age-verification wave: the EU's under-15 proposal, EDRi's critique of the eID wallet, California's new laws (the good and the bad), and Microsoft's new Windows age APIs OpenAI's rogue agents: the German-wiki hijack, 10+ more undisclosed sites, and why Henry's patience with AI companies has run out Defense Bulletin: breaches (AdaptHealth, Berlin, Trezor, Surfshark, Mathspace, Veradigm), threats (Pixel modem zero-day, Chrome, Nightmare Eclipse's exploit spree, a new Android VPN leak, LG TVs), and FOSS+ (Tor, Tails, CalyxOS, GrapheneOS RCS, Control D on Tailscale, Homebrew 7.0, and more) ๐ EasyOptOuts is the easy, affordable, and effective way to automatically remove your data from people search sites: https://easyoptouts.com Thanks to the EasyOptOuts team for sponsoring Surveillance Report! ๐งก SUPPORT TECHLORE Become a Techlorian and keep this show growing: โ Support this podcast โ ๐ SOURCES & LINKS View all sources here: https://techlore.tech/no-hack-needed-how-revolut-leaked-your-passport-plus-new-age-verification-rogue-ai-agents/ โฑ๏ธ TIMESTAMPS (00:00) - WELCOME TO SURVEILLANCE REPORT (01:04) - REVOLUT + OTHER ID LEAKS (07:36) - AGE BANS & AGE VERIFICATION (12:24) - OUR SPONSOR: EASYOPUTOUTS! (13:30) - CREEPY AI COMPANIES & CEOS (17:48) - DEFENSE BULLETIN: DATA BREACHES (20:50) - DEFENSE BULLETIN: THREATS (26:22) - DEFENSE BULLETIN: FOSS+ ๐ฅ VIDEO Watch on YouTube ๐ฐ ABOUT SURVEILLANCE REPORT Your essential weekly tech news podcast covering the latest threats to security, privacy, and digital freedom. Hosted by Henry Fisher from Techlore.
Audio and video play straight from the show's own hosting. PodSubs stores and serves no media.
Transcript
From the show. Tap a time to play from there.
This week's Surveillance Report covers the pileup of identity verification breaches from the last couple weeks, including a couple new ones, the global age verification push forcing even more ID collection. I'm sure those two stories aren't related at all. OpenAI's rogue AI agents, and the more that we learn about that, the more interesting it gets, and lots of the usual data breaches, threats, and open source updates. Welcome, everybody, to the Techlore Surveillance Report, your essential weekly tech news, delivering deep analysis on the latest threats to your security, privacy, and digital freedom, and of course, empowering you to reclaim control and defend your rights along the way.
Before we get into that highlight story, um, thank you everyone for the missed week. Uh, we just had a lot of stuff going on with Go Incognito, but it's pretty close to being ready to launch. We're just a few weeks away, which I'm really excited about, so keep an eye out for that. And also, I'll be at the next cloud conference if anyone else is going, and it happens to be in Berlin, so come say hi. I'll have these fun stickers here, uh, these like got privacy stickers. I think they look pretty cool. So if you see me, just, uh, ask for one and I'll give one to you. So we're gonna kick off this story to talk about Revolut. So for those who don't know, Revolut's a financial company. There's a couple interesting layers to this.
There's the actual breach itself, which exposed data which include people's identity, their contact details, their birth date, postal and email addresses, and phone numbers, as well as copies of their identity documents, including passports and driver's licenses. The second thing here that I really wanted to outline is how this happened, because this wasn't just your typical data breach where they didn't patch some vulnerability or they were missing some kind of security protocol. No, this was something that they just actively handed over because this came from a fraudulent request, like an email request from a government that was illegitimate because they spoofed the agency's email domain.
I think there's a lot of lessons here. It's tough because it's, it's really easy to take a valid domain as a, you know, something is safe, but there are ways to impersonate domains. So this is a situation where technology can always help if people are using DKIM across the board, which if you have your own email set up, um, you probably know what that is. Otherwise, look it up. It's called D-K-I-M. It's something I recommend everybody sets up, and they set up some protocols as well involving this because it keeps everybody safer. What we do actually internally is when we get an email and it doesn't pass DKIM or something gets flagged, even if it looks legit and it comes from the right domain, I actually flag with that team.
And this has happened with some projects behind the scenes who reach out to us for, like, sponsorships or something like that. They don't actually have their DKIM set up properly, which, A, yes, is kind of funny if when you have a privacy and security company that doesn't do it properly, but everyone can make mistakes, especially when they're early on. Um, but also it typically leads to them fixing it, and either way, um, it at least makes them have to verify they say who they are. I think the more systemic kind of takeaway that especially I saw online a lot is why is it possible for a company to just say, "Okay, yeah," like, "We get all these government requests here.
Just take this as well." So I think this is a good reminder of the kind of personal information that companies who are run by employees who are human, who make mistakes, have about you. There's this really big disconnect that people have across the board, not even with just identity verification. I think people think, "When I'm using Facebook, Facebook has my data." Yes, Facebook has your data. It's on Facebook servers, but it's also the thousands of employees that have access to your data. It's Mark Zuckerberg, like the creepiest person in the world, being able to log in and literally see what he wants about different people.
Like, I think people also lose the human element to this. We saw kind of the most creepy version of this, which was kind of those stories that started to come out that Amazon, I think even Apple and some of these other big tech companies, were using third-party contractors to analyze your voice that you were using with voice assistance, and this wasn't really disclosed whatsoever. I know it's hard to avoid financial institutions, so this is a tough one to say, "Just avoid the service outright." No matter what, try to give up as little real information as you can when it comes to these situations, right? Like, use something like a passport instead of a driver's license.
It typically has a little bit less sensitive information on there, like your home address. I would recommend everybody has a private mailbox, and ideally you have some way that you can sign up with things with alias emails, alias phone numbers, et cetera. So do your best with this. Freeze your credit. And if you have a higher threat model or a public presence, you should already have some kind of address situation so that your actual home address is kept safe. This is more of a confirmation of one that we already talked about in a recent Surveillance Report, which is that IDscan has confirmed the data breach the one that everyone knew already happened.
Um, but they did confirm that they had more than 150 million driver's licenses stolen. That confirmation took a little longer than I would like. If you missed this one, there was pretty much a mysterious website that was publishing people's IDs, and they- Saw about 150 million of them on there. It's like a dark website where you just type in someone's name, and then their ID comes up. People were speculating it came from this company called IDScan, which is used when you rent, uh, vehicles at Hertz, and also when you go to a cannabis dispensary apparently is ano- another place where they're commonly used.
Now, this investigation is still ongoing, but this is like the absolute peak example of why this random verification with random third parties with random security practices is a bad idea. I just... This makes my blood boil, guys. It's like you are handling people's most sensitive information, and to just violate people's trust like that, I think, just speaks volumes. Like, if identification companies who are handling the most sensitive data about you aren't taking security and privacy seriously, what does that tell you about people who are handling things that aren't as sensitive?
Like, this is kind of the state of the internet summed up for me and why it's so frustrating for me. And then the last story here when it comes to, like, the hidden cost of handing over your ID, which is how I'm framing this overall head- highlight story, uh, is that there was also a Florida motor vehicle database that was leaked. This came from the ShinyHunters hacking group, and it's nice because I guess two of these shortened the data breaches section this week, 'cause normally that's where the- they're in, actually. Um, but this is from the Florida Motor Vehicle Agency, FLHSMV, uh, which is a crazy acronym.
But they have confirmed a data breach in a statement last week after the hackers obtained a police officer's credentials that were stored on a personal device. That's another important lesson for both enterprise listeners and also just regular people. Don't mix work and personal when you can. This stuff can really be exploited. Now, these hackers stole hundreds of thousands of certificates of vehicle ownership records, which contained vehicle owners' names and addresses of buyers and sellers. The records also contained vehicle identification numbers, and there were a smaller number of files which did include government-issued documents, Social Security numbers, and immigration papers, but no driver's licenses.
But this is still a pretty sensitive breach. I think it's really sad because it puts a lot more on you all as individuals to have to take initiative on this. So, I mean, it really does in many ways expose the limitations of a lot of privacy advice that even I'm gonna give you guys because at the end of the day, most of you probably need to have bank accounts. And if your bank account was Revolut and you were targeted in this, unless you really, like, went through the preventative steps, which I do recommend you do, for the record, you still could have had a lot of sensitive information exposed about you. So for me, this is kind of a humbling story of, like, no matter how well we do, there are some things that are probably going to be unavoidable in our lives.
And so for me, it's a reminder of, like, yes, let's do everything we can ourselves, but we also need to be putting, you know, at least more energy into demanding better from politicians, from companies, et cetera. So that's my weird takeaway from this. Uh, but I'd love to hear what you guys think. These stories are never fun to cover. While we see these data breaches of identity verification systems, we also see governments demanding more verification through all these proposals that they're putting out.
So the EU is set to propose a ban on social media and AI cha- chats for under 15s. Children from 15 onwards can set up their own accounts under the proposal, while children 13 to 14 can ask parents to open introductory accounts for social media, and for accounts for children between 3 and 12 will be fully controlled by parents and will only be for child-friendly services with strict safety standards. And also, the thing that is a little bit more concerning to me, and this is from the EDRI, who's a very trustworthy, um, EU organization, they actually put forward why the EU age verification tool does not solve privacy concerns.
EDRI criticizes the eID Wallet, which is proposed as being very privacy respecting, but they cover all the technical limitations with it. They also say that they suggest using zero knowledge proofs, but they don't make it mandatory. So my understanding is that there is kind of the centralized open source app, um, that can be used, but every country can choose to use it or make their own. They can base it off their code, and there's few mandatory things to actually guarantee the kind of safety in every single identity verification system. And even in those scenarios with zero knowledge proofs, the EDRI still says it's not good enough.
Unlinkability can hardly be guaranteed. They talk about the shortcomings because these are not perfect technologies. We've had security and privacy scientists and researchers that have publicly commented about this. I'm not gonna get into it because it's just not the main story here. But I think if you're in the EU especially, it's important for you to read this article so that you can start developing some of your language and understanding around the tools that are being proposed to verify people, and they did a wonderful, wonderful job making this a very short... I mean, if you're watching the video, this is overall a very short article, but it's extremely dense and contains all the information you need.
It's just what you need and nothing more. Now, here in California, where I'm based out of, we are no more innocent than these other places because California governor has signed law protecting kids from risks of social media and AI chatbots. Now, this has some good and some bad, and I think the EFF does a really good job summarizing this. But there's three unique bills here that they really harp on. One of them is AB 1709, which is the bad one that we don't like. It's a functional ban on young people under 16 using social media in California. Now, two of them the EFF did support, which is AB 2071 and AB 2298, which require that children learn critical digital literacy and cybersecurity topics.
They are an affirmative and constitutional way for the state to address valid concerns about young people's internet use without violating the First Amendment rights. I think this is great. Like, education's a great approach. Microsoft has formally rolled out a new age awareness API. I don't know if it's tied be... And I, I can't, I can't directly link it to that desktop operating system age verification that is being proposed and passed already in California. Um, or it could just be something they're doing broadly because they see the pattern happening all around the world. But they have this new age range thing which, again, aligns very much with that, uh, proposal, but this is called GetUserAgeRangeAsync to retrieve age groups, while GetAgeVerificationStatusAsync can tell an app whether a user's age has already been ver- verified.
So yeah, pretty much you're getting age verification on Microsoft Windows. This is no different from what Apple's required to do on macOS and iOS now due to these laws. So this is what happens. You know, people... This is- Kind of a s- a side rant, but I, I get so many comments, uh, from, from people who say, like, "Oh, my gosh, why do you talk so much about, like, you know, reaching out to politicians and, like, standing up for things on the legal side of things? It doesn't matter anyway." Like, there's so much doomerism, but this is why. You flip it on its head, and you actually see what states can mandate, what countries can mandate, what regions can mandate.
This is what we get. We get a controlled internet. So just like laws can be used to make all of our technology even more privacy-invasive, we can have laws go the other direction, too, where they can actually help protect your privacy. I don't think anybody thinks that laws, regulation, et cetera, can actually fix problems by themselves. I don't think that's at all what the claim is, but it is a very important part of the fight. And I think the more that we appreciate that, the better off we're gonna be in the long run. So the Calyx Institute, the digital rights organization, has a little triangle which I really like, and the triangle has three things for digital rights.
It's tools, so the actual tools we use. It's using ProtonMail. It's using a nice browser. It's using a nice search engine. We have education, which hopefully is why you're here, to learn and understand what's going on and stay on top of it. And then the third thing they talk about is legal. Those three things are all very important, and there's a really tight relationship between them, so don't forget that, uh, when you're keeping up with the news because there's normally some kind of angle across all three in every single story I cover. Speaking of tools, I wanted to mention the sponsor of this podcast, Easy Optouts. Now, this isn't just your regular sponsor who just pays good money and ends up, um, on a podcast.
Like, we have had Easy Optouts on here for many years. They're just run by a couple guys. They've been shown by Consumer Reports to be one of the most effective, tied with Optery. But unlike Optery, their plans are $20 a year. There's no, like, tiered subscription model. It's meant to be extremely accessible to people, which is how I think it should be. It's already asking a lot for someone to have to pay to be removed from these sites that they never opted into using in the first place. But Easy Optouts will opt you out of all of these random people-searching websites on your behalf. You can upload as much or as little information as you want to make it easier for them to find your information, and they work all around the US.
They have business plans, so you can also get some discounts at scale. Go visit Easy Optouts and get started for $20 a year down in the description. Again, I don't think they can make this a better value. They are one of the most effective services out there. Thank you, Easy Optouts, as always, for sponsoring us, and now back to OpenAI's rogue agents. This last week, I, I don't know what happened. I don't know if it's because I skipped a week at Surveillance Report, but I have found some kind of, like, newfound rage at these AI companies, 'cause I really think this is getting to a peak.
So I, I think it makes sense why I'm starting to feel this rage, but I- I'm gonna hold it back until after I do coverage for the stories. So this came out a, a f- a couple weeks ago now, but "OpenAI agents hijacked a German website in a previously undisclosed AI breakout this spring", and "OpenAI agents repurposed the site into a message board, sharing tactics to cheat on some tasks, bypassing OpenAI's restrictions and masking their behavior." There's another Reuters article here that the rogue agents have been used at least 10 more times for unauthorized communications.
And this all, of course, came from that original Hugging Face breach, which if you remember my take, is still pretty relevant today. So, you know, they hacked Hugging Face, and the way that it was presented then was that these models just broke out. They started acting autonomously. They destroyed, you know, u- Hugging Face. It... When we learn more about the breach, it wasn't like that at all. It really did follow an overall path that a human would have followed too, just far more, you know, quick, autonomously, yes. But people really kind of overlooked the responsibility that OpenAI had in that breach, and that really is what is just pissing me off so much.
Like, the more I'm seeing these stories, the more I'm seeing people just treat the LLMs as if they are their own thing and not a tool that is managed by a company who has protocols in place to be able to control this. And when you look at these kind of responses, "OpenAI," I'm reading this from the article, "did not directly address questions about how many different sites its agents used to communicate or say why it kept the activity under wraps for months." Again, months with an M, not, not a, a Y, not yonths, months. "In a statement, it said it was undertaking a broader review of agent activity and had so far not identified other activity matching the severity or scale of Hugging Face." I, I think what's really been making me so mad the last week is these a- AI companies have proposed this vision, right?
Like, we look at Elon Musk, Sam Altman, and Dario, these three big key AI players. They are spouting every flipping day, and there was also that viral tweet recently, that, you know, "Yeah, AI's just gonna kill humanity. You know, like, we, we think it's gonna do that," blah, blah, blah. "Meanwhile, we're building those tools that we think are gonna kill humanity," which doesn't make any sense, "and by the way, we're just totally chill with them doing this, and we're not gonna actually apply proper protocols." The amount of incentives that these people have to push this narrative is too high for me.
When they push this narrative, it puts everybody in a fear-based state. It also, funny enough, puts them in control of, "Well, we have the technology, and so we need to decide what to do with it," which is what we're seeing now. We're starting to see talks about halting AI development and all these different things right now, and who's leading all those discussions? The AI companies who caused all the problems to get here. Why does that make any sense to anybody? They have done the same thing for years. Many of these people literally are, like, narcissistic assholes that just have very low bars for ethics.
They run these companies that are move fast and break things. They're fine to exploit you for years. They're fine to leave your passwords in plain text. They're fine to leave all their customers/you, at users, exposed, and they don't care about your safety. They don't care about your privacy. They don't care about your rights online. Facebook was literally a core part of a genocide in Myanmar. That is something that happened. We also see how they're fine exploiting children even though they know- Full well that what they're doing harms teens' mental health. Like, these are well-established things that these companies are fine doing.
So when we actually come together as a society, when politicians say, "Hey, let's ask the AI companies run by these people," who I think are kind of crazy, "Let's ask them what they think we need to do to rein in this technology," that is the wrong discussion to be having. And then you might ask, "Okay, so who should be running that?" And that is the problem. We don't have anyone right now to hold these people accountable. It's very scary because a lot of times these companies are writing the laws that your politicians are signing. So I don't know where my rant's going. I feel like I got really rambly there.
But I, I, I've just been really frustrated by this. I would love to hear your guys' thoughts in the comments. All right. Now we have the Defense Bulletin. Again, three sections. We have the data breaches, we have the threats, so you know what to watch out for, and we also have the open source updates. Um, and I'm gonna be a bit quicker this week because I'm looking here and it- there's probably at least, like, 30 to 50 stories across these three sections. So Adapt Health confirmed a 4.1 million data breach in a July cyber attack. This is your reminder that breaches happen typically retroactively.
It's quite rare for a breach to happen, and then you just find out about it right when it happens. So if you have any kind of relationship with this company, check out the show notes. Um, this one was a Berlin data leak that ended up on the dark web, and it really calls into question Germany's IT infrastructure. And this is from TUTA, and they talk about what happened and kind of their take about it. So especially if you live in Berlin and you wanna learn more about this, I would check this out. Trezor, the cryptocurrency hardware wallet, had a data breach that impacted about 81,000 customers, which is, I guess, currently what the tally is.
It includes full names, shipping addresses, email addresses, and phone numbers. I'm really not a fan of these data breaches because it exposes potential customers who are holding potentially a lot of money, and it puts a huge target on their back. And so if you purchased a Trezor, you should look to see if you are impacted by this. And if you were, I would just take appropriate precautions from there. On this note, just a heads-up, Trezor has warned people of phishing attacks probably as a result of that breach now that they have people's emails, so just be extra cautious of those kind of emails. Surfshark VPN has said that hackers breached their internal testing proxy servers.
Um, they did say it didn't affect customers and it didn't extend to other parts of the infrastructure, but it exposed service con- uh, configurations and build-related credentials. So it seems to be a pretty small-scale breach, but if you wanna learn more about what happened there, again, there's always the show notes. By the way, on the show notes, uh, like, about every week I get a person saying, "You know, you don't put the links in the show notes." Uh, the way it works, there's one link in the description. It's, it's linked. It says, you know, "Click here for show notes." You click that link, and guys, remember, Surveillance Support is a newsletter and a video and an audio podcast, so it's kind of hard to do everything for everybody perfectly.
But if you click that link, the very top of the link, it says, "Click here for all sources." Then you just click that, and it'll take you to a dropdown that has all the sources there. The reason it's set up this way is because that page that you click on is also sent to people as a free newsletter each week, and so that's why the sources are on the bottom of that page, so I don't just spam people with sources at the very top. And that's also a way for me to say that you can subscribe to our newsletter and you get, like, a five-min-minute written version of this podcast if you prefer that kind of thing. Now, a French hospital was fined โฌ500,000 after a breach exposed data of over 700,000 people.
So this is probably for a breach that I covered someday in the past, I'm assuming. Um, but this was just more of a, like, consequence to that breach. Now, there's something called Mathspace which disclosed a data breach affecting over a million people. This is an online math learning platform, so if you use this, please, again, check out the show notes. We also have Veradigm, I think that's how you say that, that warns of patient data breach after a ransomware gang claims attack. This is a health tech company, and it's a third-party vendor that exposed all that information. All right, now we're gonna get into the threats. The first one is that Google says some Pixel phone owners were hacked in zero-day attacks.
This is from a bug in the Pixel phone's modem, which lets the device connect to the internet. So exploiting the bug could allow an attacker to gain access beyond a sandboxed wall of the modem and into the broader s- phone data. The concerning thing is I still haven't seen any kind of information about this, like how many people it impacted, who was behind it. As Zach Whitaker says here in this article, this is typically used by spyware makers, and so that's also where my mind would've gone. But again, there's no proof of that yet, and there's no proof of the scale. Now, it has been patched, so this is no longer something to be concerned about.
But I would have loved to get more information, and I hope that Google at some point does come forward with that info. This was a really big week in terms of Chrome vulnerabilities. They had 230 vulnerabilities on Tuesday. So you should always be updating your browsers. They're one of your most sensitive pieces of software, so make sure if this is your reminder, update your browser right now if you haven't already. Kind of on this note, Chrome is now shipping updates every two weeks as AI changes the security landscape. That's how TechCrunch, uh, kind of frames it. Uh, Mozilla, Microsoft, and Brave have also said that they're, uh, adopting a faster two-week, uh, schedule following Chrome's lead as well.
So this is about to be, uh, pretty much a universal browser thing at this point, except for Safari, who has their own little update, uh, system that doesn't align with anybody else's way. Uh, Mr. Nightmare Eclipse is still here, man. I- every time I cover a story, I'm like, "You know what? I'm seeing less of these Nightmare Eclipse stories." But he comes back. So Nightmare Eclipse is a disgruntled security researcher, um, who has released a CrowdStrike Falcon zero-day exploit named FalconFlank. And he also released privilege escalation zero-days for Kaspersky antivirus, for Endpoint and Gen Digital Avast antivirus, as well as a denial of service zero-day for Nvidia that will crash the system, and so they're, they're on a roll, man.
I don't know how they're finding all these different zero-days, but Nightmare Eclipse is, is really... Yeah, they're, they're doing it to them. One of those, if you wanna learn more specifically, is called ShieldCrash, which impacts Microsoft Defender. So again, they're still going after Microsoft. I... Man, I'd, I'd love to, like, see an interview with this person somehow. I'd love to hear, like, how they're doing it, and also just... Yeah, I just wanna learn more, man. This one is something that I've been having to say a lot in content, which is that we just can't trust mobile VPNs as much, unfortunately.
And iOS and Apple have actually taken a lot of flak for this the last several weeks, but Mullvad just came forward with this. Another way to leak traffic on Android has been discovered. And this is even if you use that setting, which is block connections without VPN. So the leak involves telling Android to create a keep-alive UDP connection that is, uh, offloaded to the hardware Wi-Fi or cellular chip. A malicious app can misuse that to send UDP packets on port 4500 to any server on the internet. So this does require you to have a malicious app to misuse this, but the fact that that, you know, capability exists is obviously something to close up.
So the real conclusion, and Mullvad says this, use trusted apps on your device, and hopefully we'll see patches coming soon. Graphene has said they are working on a fix, but it doesn't look like it's live yet. So right now, just install trusted apps. This one, if you are on YouTube, have-- You've probably seen this more than enough times. The coverage for this has been crazy, and I, I missed the wrong week 'cause this would've been a good highlight story for me. But LG TV was shown scanning your LAN for third-party phones and other devices. And this really came from GamersNexus on YouTube, who worked with Level 1 Techs, which is a tech-focused YouTube channel.
And they analyzed, uh, LG TV's tracking capabilities. They shared packet captures that pointed to the TVs, um, and it pretty much, like, literally tries to find everything. The concept and the concern here is it collects this data even if the TV is offline. So if you have an LG TV, it's just constantly, constantly, constantly collecting all of this data, and the big fear is that it either connects to some kind of, uh, public Wi-Fi network, or if you ever connect it online someday, it actually will upload all that data retroactively because it's not actually that hard to store that data for a long term.
I actually think one of the most timeless videos I made was my video on, like, why I hate smart TVs, and so I'll leave that linked as well if you wanna see that. Um, but it, it just, it covers all of these concerns I have with smart TVs from, like, five years ago probably. And it's crazy to see, like, real information now really expose how, uh, kind of exploitive this can be. Now, LG does downplay this. They say, "Oh, all smart TVs have this. This is native functionality." My whole counterpoint to all of this is why not make it easy for people to say, "No, I don't want this.
Just don't collect that data"? If you're not doing that, I'm not gonna trust really much of what you have to say. It's not that hard to say, "Hey, yeah, we're gonna collect all this invasive crap about you, but we're gonna tell you, A, and B, uh, here's how you opt out." Not a hard thing to request. That should be a bare minimum standard, and in fact, I would say there should probably be some kind of laws or legal requirements for companies to have to do that. That is not a thing in the US. Moving on, we have 36,000 exposed Plex servers vulnerable to recent flaws, so if you have a Plex server, make sure to check out that story.
Skullcandy Dime 3 earbuds expose users to Bluetooth hijacking, so if you happen to have one of these or you're just curious to learn about this, 'cause this is an interesting one, I would check out the show notes. There was a cPanel backup plugin called Acronis, Acronis, which, uh, yeah, it, it has an exploitable flaw. So make sure to check that out if you use that plugin. And on that topic of plugins, we have WordPress plugin issues. One of them is the WooCommerce plugin. The other one is Admin Menu Editor Pro. That, that is the name of the plugin.
So if you're a WordPress admin, make sure to look into those stories and get your stuff patched. All right, and now the open source section. Tor Browser hit version 15.0.22. Your reminder to be updated. It seems like it mostly includes security updates. Tails, the anonymous operating system, hit version 7.13, which updates brows- uh, the Tor Browser to, um, the newest version. It updates the Tor client. It simplifies the shutdown procedure. Um, and there was also a few other changes there. Now, there was a security update from CalyxOS, 7.2.5.20, uh, which includes some new stuff, uh, but it seems like it's mostly security there.
Uh, Graphene OS is pretty exciting, has planned to overhaul the bundled, uh, Android open source apps, including messaging with RCS support. The reason why this is big is that we finally have seen end-to-end encryption rolled out in the RCS protocol. So this is meant to be a better version of SMS, which means anybody with any cell phone can communicate with each other over their cell company. But up until now, RCS, uh, especially with end-to-end encryption, was only found in Google Messages, uh, for the most part. And so what I'm really excited to see is some kind of RCS first party with end-to-end encryption, and if Graphene can pull it off, that would be awesome, and hopefully they keep it open source so other people can also copy that too.
Uh, Addy.io, which is a really nice email aliasing service, they're fully open source, and they let you keep your email safe from countless websites, now blocks email tracking pixels by default. So if you use Addy.io, that's a nice little change there. They've also partnered with something called Leave Me Alone, uh, which is a privacy-first tool to unsubscribe from newsletters and manage your inbox. I have never heard of this service, uh, so I'm not, like, formally condoning it. But if you are an Addy.io, uh, customer, you get some kind of discounts here, uh, with Leave Me Alone as well. Now, image hit version 3.2.
This is an open source photo, um- Tool, service, uh, from Fudo that is typically self-hosted. Um, it debuts a new search API, cluster groups, workflows expansion, and much more. So if you use Image, uh, check out the f- the show notes for the full changelog. This one really flew under my radar, but Control D is, uh, there's actually a couple things here. So the first one's just the story. Control D is now available through Tailscale. Control D is a DNS service from the team from Win- Windscribe, actually. It's pretty much NextDNS if you guys have used them.
It's a way to control your whole DNS. You choose what filters you wanna use. And now they're supported in Tailscale, which is a VPN. Not like a traditional privacy VPN necessarily. Um, Tailscale lets you connect to, like, remote devices. So, like, I have my NAS behind me, and the way I connect to my NAS remotely is through Tailscale. Now, Tailscale has done a partnership with Mullvad, so as part of that same VPN connection, you get routed through a Mullvad exit server, so you get Mullvad's protections, and you can connect to your NAS, which is why I use Tailscale. But now you can also use Control D as the DNS provider by just pasting your ID from Control D directly in there, and then you'll start using, uh, Control D automatically.
It's pretty cool. You could have always probably set Control D manually, uh, in this, but this is just a nice partnership that makes it a little bit simpler. Now, the second thing I wanted to mention here, this isn't new whatsoever, but it's new to my brain, uh, because Mullvad actually shut down their encrypted DNS resolver service. Well, they haven't shut it down yet, but they're going to, and this is, comes with out-of-the-box filtering that's maintained for you, and it's something I've been using for a long time. And I just wanted to flag because I covered this news last week. It was really sad, and a lot of you expressed, um, that you are also sad about this.
Um, Control D actually has the same exact thing, and this is the perfect swap-in, and it's what I started using. So you can see here if you're watching the video, they have a malware protection. That's just malware filtering. They have ads and tracking filtering. They have social filtering. They have family-friendly filtering, and they have advanced. This is completely free. You just click Select, and it tells you, um, if you wanna use DoQ, if you wanna use HTTPS 3. Like, you can use whatever you want for free. You don't even need an account to use those. You don't get the customization, um, but I think this is the best kind of swap-in, uh, for Mullvad's DNS service that was recently announced to be shut down soon.
So just a heads-up, uh, because I thought this really solves at least my concern. This is a quick PSA that Switzerland's federal government is testing open source alternatives to Microsoft on 3,000 computers, so I hope that's successful for them, and it's exciting to see countries really test that kind of stuff out. Android is rolling out passkey transfers between password managers, which is native to, uh, the operating system, which is pretty exciting. We also have Homebrew 7.0, which is a package manager for macOS, which arrives with faster installs, hardened security, and a native macOS app. The native app looks really clean just based on the screenshot, and it's nice to see security get better.
Ente Photos is open source end-to-end encrypted photos. It is typically hosted, but they also have a one-command self-hosting option as well. But their machine learning is now faster, so if you use that, that is good. Uh, pretty cool. They also have introduced library sharing, so you can share a whole library with a partner, which is pretty awesome. And now they also have a service that I covered, uh, I think a couple months ago now called Locker, which is a separate app, uh, from Ente as well, which can now scan paper documents. So you can keep documents in there, uh, based on a scan.
Waterfox, the open source Firefox fork, hit version 6.7.2, which is more improvements to their recent big update, including improvements to tree tabs and the vertical sidebars. So I'd say this is just overall polishing up their recent pretty huge update for them. I'm excited for Waterfox. They've had some really good stuff. And also they released an Android version, 1.2.7, which has tab groups, homepage improvements, and privacy controls as well. Now, open source X front ends Nitter and XCancel resume their services after seeking legal advice. So X, formerly known as Twitter, uh, sent a cease and desist to these open source front ends, uh, to allow you to access fricking posts without needing an account.
It's like, oh my gosh, I'm a social media company. I'm gonna make it harder for people to access content by account gatekeeping, account keeping literally as much as I can, and then we're gonna get open source front ends that will make that information publicly accessible anyway. Oh, now I'm gonna send them a cease and desist. Yes, I'm the good guy in this equation. They haven't really commented on much, but feel free to keep using them. This came from Brave, but Brave did say that their desktop outperforms other browsers in both speed and performance, and they put all the numbers together. They found they use on average 44% less CPU, 10% less energy, 28% less memory, pages load 20% faster, and also faster transfers for inbound and outbound data.
They go through the benchmark. They went through how they do this and resource usage. So i-i-it's pretty compelling stuff, and I think it really speaks to really the footprint of tracking online 'cause that's really the main thing that Brave is probably shutting off here. They're shutting off ads. They're shutting off trackers. So yeah, pretty cool stuff and one more reason to use any kind of browser that's gonna help you there. I love stuff like this. I love benchmarks. I love data, so it's pretty cool to see it actually measured. This is a heads-up that Surfshark has announced their acquisition of data removal service Optury.
So Surfshark is actually owned by NordVPN. A lot of people don't know this, but NordVPN acquired Surfshark. Surfshark has this service called Incogni, which is a data removal service, and now Surfshark has acquired Optury, which is a data removal service. So, you know, a lot of messiness there. I see this as kind of the same consolidation we've seen in the VPN industry. If you've kind of missed this, most VPNs you see out there are owned by just a few of the same companies. Y- we're starting to see more of that consolidation now, it seems like, in the data removal space.
So yeah. Shameless plug, I mean, I don't have to promote our sponsor outside of the sponsor segment, but seriously, easy opt-outs, uh, independent, just a couple guys, like very humble dudes. It's $20 a year, um, and I, I don't see them doing anything like this anytime soon. Debian hit version 13.7, which updates Trixie with security and bug fixes. And if you wanna see the full changelog, check out the show notes. Plasma 6.8 KDE, um, Beta has added CUP backups and expands Union theming, so that's just a heads-up there. CUP is a backup scheduler module.
Um, so yeah, uh, it should be exciting. Uh, and also we have Linux Mint who's unveiled a new e-book reader, a calendar app, and a revamped document library, so some updates there for any of you who use Linux Mint. And Asahi Linux has added expert mode support for N3 Macs. This release includes webcams, microphones, Wi-Fi, Bluetooth, USB 3 at up to 10 gigabits per second, and hardware-accelerated video decoding with AV1 support. I love Asahi Linux. It is such a cool project. It allows you to install Linux on any kind of Apple Silicon devices.
At least, I think it sounds like M1 through M3, um, are supported And it just keeps getting better guys. Like I was testing it out on my last MacBook Pro and I really liked it. I actually think that the MacBook Pro is like the best Linux machine out there. Maybe even the MacBook Air or the Pro, like either one. Uh, it's really nice hardware when you combine it with something like Asahi Linux, it was just so much fun to run. And it made me realize how behind Linux is in the ARM world. So, uh, kind of, uh, pros and cons, but overall I'm just hyped to see this. All right, everybody. That was a bit of a longer week, but I wanna thank you for getting through the two weeks of news.
If this analysis helped you out, seriously, like the way it's possible is through so much of your guys' community support. So I have a private Signal group that I've set up just for all of you who wanna join. You can become a Techlorian. It's $5 a month. You join that Signal group, you get access to other people there. You also get access to my private RSS feed so you can keep up with these news as they go live. And also you get to support this content. And there's higher tiers as well for Techlorians. Like I have a monthly meeting one where you get to chat with me, and it really helps us out, and it's even cheaper if you go annual too. If you can't or don't want to support via the paid avenues, um, you can always support for free by just giving this a like on YouTube, sharing it around with people you know, and leaving a rating.
I do see your ratings, and they also really help, um, on both Apple, Spotify, really any podcasting platform. So if you're able to leave a rating, please, please, please do. Thank you all for listening. I'll see you in the next episode of Surveillance Report, and potentially in Berlin if any of you are gonna be there. And I'll see you next week.
Notion's AI chat transcripts were exposed through its AI provider, Fireworks AI, and "zero data retention" turned out to underperform. I also cover Quad9 pulling out of France over site-blocking orders, 333 hijacked Tor...
Congress now has two site-blocking bills, and one of them would force VPNs and encrypted DNS providers to block websites for Americans. Plus: ShinyHunters hacked the FBI and someone stole 3 million Pentagon personnel...
An ID-verification company, IDScan.net, appears to have leaked scans of 153 million driver's licenses, then sold on a dark-web market called Nexus. This week I break down why this breach is the perfect example of...
Four positive VPN stories in one week: Mullvad's new multihop modes, Obscura shipping on Windows, Apple quietly patching the iCloud Private Relay leak, and Firefox's free VPN rolling out to Android. Then it gets darker:...
Three cracks opened up in Apple's privacy story this week: iCloud Private Relay leaking real IP addresses through DNS prefetching, WebTransport, and passkeys; Hide My Email exposing real addresses in bounce logs; and an...