文字稿 🔗
Anna Vanderberg: Hello everyone and welcome back to another MSS edition of Initial Access. If this week gave you shell shock, you are not alone. two over three stories to cover this week. and in in shell, quite literally. And I will get better with the puns and the jokes, right? that didn't sound that that good, I I think. they the the thing about these two per particular vulnerabilities is that they hit two of the most common runtimes in in the internet. And on the same week, right? It was really crazy. Again, today we are looking at this from the mana service side, how so something in particular, like or things in particular to highlight how fast the exploitation timeline move, how hard was to like find every piece that was vulnerable in in in both of these vulnerabilities. And yeah, and we will kind of briefly talk about a new ransomware, a new group that's just going for the big names. my name is Sergio Villegas. I'm a senior money analyst here at Visual Fox. And today I'm joined by Dylan Sparks, senior operator, Emilio Gallegos, adversarial operator, and John Ons, senior security engineer. Welcome guys. it's always nice to have you. how do you feel about October and all of its vulnerabilities? I mean it's definitely a spooky season, right? It's funny because it's like nothing ever comes just like one at a time. Right? like i it any any any sort of bad event, it's just all at once. All as much as you can handle, and then we'll have it. Maybe, you know, maybe this is, you know, knock on wood. Maybe this means we'll have a quiet, you know, rest of the holiday season. Just get all the vones out now, right? Yeah, I don't think that will happen. But it's it's it's it's look a man can dream, okay? I'm there with you. so why don't we jump directly to tour stories, right? Like There are really cool stories. they are on the on everyone on every every news cybersecurity outlet that I'm seeing, every IT outlet I'm seeing these vulnerabilities. Even on Instagram. I saw like the other day. My my Instagram feed is not like cybersecurity stuff. It's a very normal Instagram feed. And I saw reels about this vulnerability and how and why it's really bad and why it's like impacting all over the place and and why everyone is screaming. Right. So so let's talk about pitch peak to shell. And this will be a recap one quickly. I know for for those that listen us on a regular basis. I know we talked about this last week, quite literally. so PHP to shell, kind of a critical, very critical vulnerability because of two critical flaws on the PHP engine by itself, its two vulnerabilities, CV twenty twenty-six ten ninety-six eighteen and ten ninety-six nineteen. both With a nine 9.2 Cv C VSS score. long story short, what this allows is for an attack to to perform remote code execution, right? If you know conditions and whatnot. But it's it's fairly easy to exploit this vulnerability. the whole thing about this and and things that I want to bring to the table to discuss is
Load full transcript
one is the timing, right? The timing of this vulnerability was very weird because It was supposed to be released later last week, but because of some tweets that we saw, it was like an overnight situation where t all all of the news media covering vulnerabilities like exploded, right? Like hey, PHP. It's it's kind of a very critical situation. So kind of question for you guys, right? Like, how do you handle like having your normal vulnerability bulletins, right? That you have your Patch Tuesday with Microsoft releasing vulnerabilities on a normal cadence and then you saw and and then you see a random post on a tweet saying hey, here is something and it explodes, right? who who what what's the difference? Is is there any virtual difference between tweeting those two scenarios? Or or it's the same thing, right? Yeah, I think so yeah, like you said, this was like a a weirdly unique well, yeah, kinda unique one because it was in an it it was it was dropped ahead of time in a lot of places, right? There was I think it was Wee's was it Wiz their their their video that they put out on Twitter and then like Emily had to pull it back because you know they b think they they were one of the parts of the I guess the the squeaky wheel. but like I I first saw it on I think it was on Hacker News, like Y Combinator. just a random post. Literally just a singular post that was just like, hey, by the way, two to like two 10.0s are about to drop. from that it's like it it I I don't know necessarily that changed like the the work approach, but it certainly changed like the criticality of it, I think. Because now you know that something's coming down the pipe and you have no information at all to go off of, right? A a Vuln in a language is like there you this is an actual needle in a haystack with nothing to go off of. We hadn't have any you know what I mean? Like there's no patch to compare against. You're you're just and you don't know the class of bug, the vault the the criticality of it. But other than that, I there was no place to even start looking. at least from like a yeah, good. I think that's a really good thing to hit on too, is just like how vague the initial like signal flares were. but in the day of like AI and you know, we'll call them self-starting security researchers, I think that also just like negatively contributed to the amount of like distractions that are out there. when you like vaguely are like, man, there's a a serious vulnerability in PHP and tweet. Right. Awesome. Great. Like there's there's nothing else out there to like help legitimate security researchers weed out the noise either. Right. Like So it's catch 22. Anyone can grab PHP, throw it into a language model, and hopefully churn out some code that they don't understand. And that's that becomes part of the noise. And I think that also leads to the confusion about like we we've talked about it here before. Like, what's a legitimate exploit? And what is just an exploit attempt by someone who's throwing code and doesn't know what they're doing? Right. So I think we saw it too. People were like, Well, we broke the embargo or whatever because we saw we saw why that's fluids. And then it's like, well, legitimate security researchers are looking and they're like, this would never work. This isn't even targeting the actual vulnerability. Yeah, yeah. That's that's a that's a good point too, right? The the immediate I think within the first, you know, two two days before the the actual patch got released, we were seeing all sorts of just like absolute just dog water code being put out as far as scanners and you know, proof of concept exploits that I like, I don't even know what some of them were trying to do. And in addition to like malicious code that was being put out there where they were, you know, saying, here's a good pock for you, you know, this works on all these different versions. And it was actually just like a self-pwned script. you know, and then and and I think it even got worse after the patch came out because now there's there's there was like an an inherent kind of confidence at that point that a lot of people gave to a lot of the the scripts that were being put on post post patch because it's like well you now we have the patches we can we can compare them. Clearly this other person was probably doing the same thing and and you know that that wasn't the case for a lot of them. And and that makes it interesting, right? Because not that you have a patch because the timeframe was like between the tweet and It it was seventeen hours or of of absolute garbage of i between POCs, malicious code, lit yeah, quite literally seventeen hours of that. And then you got the official release and suddenly every POC is clean now, right? okay, now you have but something I noticed is that the majority of them were like the same thing. As you mentioned, right? yeah because LMs are so good, now everyone has clean code because they are comparing the the the right version changes between the the engines and it's like Yeah it's it's crazy, right? would you like you you guys dealing with this every day, dealing with having to manage vulnerabilities every day across entire attack surfaces. Question is, it what do you prefer, right? doing it like every two every patch Tuesday, having a hundred vulnerabilities with a high level with a lot of documentation where you can actually act or having a random tweet, right? Do have a prefer a preferred way to handle it? Or it doesn't matter. I I I I prefer having all of the information directly given to me. Absolutely. It makes my job a whole lot easier if somebody tells me point at this line in the code, right? Yeah, yeah. the more information that's out in an advisory, I i I think it's twofold, right? I think for me as a professional, yeah, of course it's gonna make my job so much easier if an advisory literally dictates to me the patches and this function and the you know this service and we fixed
it in this way. but also on the flip side for like for vendors, I think it like we we've seen this problem before of like when you put out something nebulous like like in this case of this phone, you have like a number of other researchers or sorry, a number of researchers that all just kind of dive on the problem because it's an open it's an open puzzle. But if you give the solution to the puzzle, like right at the right at the start, you're gonna get a lot a lot less of that like incidental attention, I guess would be the best way to describe it. so like it's a catch twenty-two there a little bit because like it's always good to have more eyes looking at at code, right? Like I'm I'm all for having as you know as many people that have spare cycles to to to give looking at new volunes or or looking at doing doing zero day research and and just improving security in general. but at the same time, I understand from like the business side of things for a vendor, they they probably don't like it whenever, you know, any any of us researchers just come in or like, hey, by the way, we found these 17 volans, you should probably patch them today. And there's probably gonna be a lot less of that if like if if the the actual if the the Vulan that they have released is so detailed as far as the advisory disclosure and all that. I w I would just like to mention that the way it happened for me or the timeline for me, I started seeing a lot of negative reception, not just for the quote unquote researchers that were coming up with their own POCs and breaking the internet, you know, with tons of views and linking them to their own repos and et cetera. But I also saw that a lot of people were sort of Commenting on PHP being irresponsible for the way they handle things. But to me, it just it sounds like they don't really know how much effort it takes into backporting fixes and software like PHP. Because at the end of the day, they presented it, these two bones with no details, but they did say that the patch was coming, right? So but I I don't remember who I was talking to that It felt like this is now an open invitation for everyone in their AIs, right? If you're saying that there's a patch coming for your thing, it just means, hey, look at my thing with all of your agents and see what's up, right? and I definitely don't feel like this is just gonna be like a a one-off incident. I feel like this is going to force us to move around this sort of new I guess we can even say that there's like a new entryway for different individuals with AIs that want to sort of test the waters with cyber and and do their own thing. But the more we saw all of these POCs and the more testing that we did internally, the more we saw that given the way the vulnerabilities work, it's so dependent on you know server configurations. And it was even dependent on like your own CPU like specific configurations as well. So It's it definitely worries me, but at same time it's kind of like a new challenge. I don't know. Yeah. Well and like and you said, right, the the configuration matters a lot with this one. I think I think understandably, right, a lot of the initial reporting about this was like, you know, this is such a huge widespread vawn and everybody's vulnerable and patch now and like those are all technically true, right? I'm never gonna tell an organization to not patch for a vault. but it you know, exploitability is like is a big deal. you know, th this vaune in particular, like we said, because it's it's something that's in I mean, this is like even more foundational from like a like a code perspective than like we we see a lot of stuff like libraries. Like this is in a in in PHP, like this is a lot more down in the weeds when it comes to how you have to patch it. Because a lot of times these are the you you install PHP as a dependency for whatever other application you're running. and so I think that's like a big thing too is like with embedded technology and and like people targeting now like embedded languages of code. Yeah. It it makes fingerprinting and identification significantly more difficult too. Yes. Because it can run in a myriad of different applications, software. And now you have to look at your inventory and fingerprint or create detections for several different tech stacks that may or may not have PHP in it. And even then, some of those don't like to play well with version headers or anything like that. Like some of these things just don't report out. versions externally. So it's like you have to have that internal knowledge base. It's becoming more and more important that we have like more mature security postures internally and practices than ever before. Like I think asset categorization and asset tracking and management practices are of like the utmost importance now, like more than ever. just because like I think we're gonna see a lot more of this stuff. Yeah. Yeah, I agree. And and I think this is a a a great segue to our second story, right? Which is I'll say the same, right? and we're talking about note to shell, which is quite literally another embedded kind of situation where we have a lot of packages using nodes and particularly and This per this story in particular, right, just to recap a bit about it, it's similar and out of bound right in the Google's broadly decompression library, which is kind of interesting because the vulnerability by itself it's not about node in reality. It's about the broadly library which just happens to exist in node, right? the th the thing with this vulnerability is that node is very famous and that's why it was dubbed node to shell. I like By the way, we were talking earlier, and I like the the name we are giving to this vulnerability, which is properly to shell. I think that's pretty cool. and it it's it's a similar situation. We have our potential remote code execution that you can leverage through an out of bounds write in this library. it it is a 9.2
vulnerability in the it was found similar situation. It was found by OpenAI codex. It's similar in in in the sense of AI, right? PHP to shell, it was discovered because of a bug that was reported several years ago. And then someone decided to scrape again with their with their AI and just happened to find the perfect scenario to do a memory corrupt corruption and exploit it basically. And this is kind of the the same thing, right? Like this was this was a vulnerability that was found because AI decided to look after like comparing versions. And by comparing versions, what I mean is the the eight month gap that we have. There was a fix available in in Google's Broadly repository in February twenty seven, this year. There was no release, no advisor, nothing. There literally the the fix was sitting there. But it wasn't patched in the official release until OpenAI decided to uncover this vulnerability. And now we have the this fix, right? similar situation, no it's available everywhere on the internet, right? Sorry, it's one of the major engines that you could be running on your on your web server. It's very popular. It's embedded across many things. Similar to like if you have WordPress, highly likely you have PHP, highly likely you have a plugin that is pulling node resources, right? So you could be vulnerable from both sides just by having WordPress installed in your environment. So similar situation, right? Like and and and here the question is. Okay, we know there is vulnerabilities happening, we know there is a lot of information on the public. This this wasn't tweeted, th there wasn't a tweet out there with a video. That's that's good, I guess. But there was information disclosed on the public for everyone to read. How we deal with that? How we deal with is is this is this a zero day? Well not really. It's been there since February, so you could argue it's not a zero day, right? But how do you deal with like Now we have to look at every repository of every major framework and look for all of the old stuff to see if there is any bug, right? we are already dealing with today's vulnerabilities. How do we leverage the time? And this is about how we as people that manage our customers' assets and we try to protect them. How do we handle and juggle the time between doing that, protecting them, and looking at these major frameworks of maybe there is something? Right, it's it's hard, right? It's it's not easy. Well it's yeah, it's a scaling problem for sure. Yeah. no, go ahead, Don. I think you've you've been looking at this one a lot too. Yeah. Like monitoring monitoring upstream commits on everything in your tech stack. Who wants that? Not feasible. Come on, man. You're not doing that. You're not doing that. Sick Emmy on it. Yeah, all right. I mean, you got it. yeah. Would you like to come to tea? for for context for the people to listen to us, Pinon in particular, he works on the on the threatenable men team. What they do basically is that looking
at the tech stack of all of the major software across the world, across the universe basically, and looking for like potential vulnerabilities and and releases, right? To see if there is anything important on those vulnerabilities. And then match against all of our customers infrastructure. So they're basically yeah, they are like Saron's tower looking at everything. That's like yeah. Yeah. Yeah. We're tired. That's what we are. We are tired. We are tired. Yeah. This month has been the longest year of my life. Absolutely. That is a hundred percent accurate. But yeah, yeah. Go ahead. No, no, no. You mentioned it, right? Like spooky season. It's it's just it is spooky season, yeah. No, I think like like this is a good case study. Like you I I liked how you framed this, just comparing the two the the two big to shell bones because I think that's like that's like a solid way to look at it, is The first question we we we kind of like approached this with was, you know, should should we or should we not, you know, or or rather what is the difference between like, you know, getting notice about something ahead of time versus just like boom, here's a C V E? and like these are th this is like the exact right case study to to kind of pull between those two because they're both I don't I don't I I can't recall if it's like a logic bug or if it's a memory bug for for Broccoli. I think it's a logic bug. But they're both it's the same tech stack like placement, right? These are both, library level, language level bugs that are affecting a number of different products. It's just one had a you know, couple of days head start for researchers, and the other one, well yeah, it's it was sitting there in the code base. I think the original I gotta go find the the actual commit because I from memory, like the the the commit that introduced this Vuln. was to fix another vault. it was some some some semblance of that of like it it was like a p it ended up being like it's yeah, it fixed something, but it introduced this into the into the stack. which I like I I always like those are so frustrating. Like for from like a developer standpoint, right? You who spent all this time like doing all this debugging and then like yes, solid, gotta fix. and now you've introduced a critical vault. I feel bad for that whoever Who yeah, right? It's bad day. just feel feel real bad for whoever whoever committed that particular that particular branch. but yeah, I think I don't know, like this one this one's a little like I feel like this was one that like it came off the back of PHP DeShell. and it it it was just an easy one to like latch onto. It was Hey, here's the next big new thing, you know, let's you know Let's keep this momentum going. Let's get big phones out there. and it and again, it is a big deal, but I've not really seen much in the way of exploitability outside of like I know the like the node, like raw node.js, that was shown like pretty pretty quickly. Like there was there
was a proof of concept out for that. other other platforms and products that employ the Brotley decoder, I think they're like I I think there's been less attention to those, but I'm not I'm not as familiar. Like I'm I'm I'm up front on this one. I haven't looked nearly as much into the Bratley side of this. Just because yeah. There's at least three major ones. the.net Visual Studio, Over Shell and P and Python. But they those already got updated because they do use the the Bratley Did they have did you have you seen Or has has anybody seen like any amount of reporting like of hey the we proved exploitation with like the Python implementation or with the the dot net implementation? Not really. And I think it has to do with l which is an interesting question, right? The the scoping because how do you scope for and and similar to the PHP, right? Like you are not scoping for the for the framework for the for your Godpress, your front face, you're scoping for something that is embedded, right? and this is a similar situation. You're scoping, how do you scope for something that is embedded into so many things, right? I think it's it's also embedded into VMware related stuff. So and and Dylan, you you mentioned it, right? no more than ever you need to do like you have a consistent inventory of everything that you own. So whenever something like this pops, right? Like you know you have it and you know you can do something. But like Doing an inventory of embedded things, yes, I have broadly inside my notes. You you don't do that, right? it's it's very tricky. And I I I think if someone does it, kudos to you. But I don't think anyone is really trying to it would be it would be hard to track 'cause a lot of these sounds like yeah, in the in this case, I think I think that would be the the the the broadly decoder aspect of it, maybe that would be a piece of code that gets updated when you do it like Update a specific application that's using like it's bringing along its own whatever.NET decoder or or Python implementation or whatever. But but a lot of times those are just like prerequisite installs, right? Like same thing I have with BHP, right? I it just says you have to install at minimum this version of node, Python, whatever. Yep. and so yeah, I I ag I agree. I I do like a hundred percent agree, right? that that sysadmins, Sysos, all of that whole set stack of the company needs to be like as informed as possible about what's running within their infrastructure, but it's such a big lift to like to track all that down because you know knowing knowing like the ground source truth of like well I did update what but what updated. Right. like look at what is it? The what's the like apt repo? Like for for any of these? Like how how often do like does the apt repo lag behind on security patches because like for whatever reason, like it's not intentional, it's just like that's just how those repositories work. So if you're not going and like grabbing the source code and compiling it for your system, are you sure you're up to date? Like, you know what I mean? One percent. that that's a great example. I do manage a couple of Debian servers that I have myself, and for example, for the PHP one. So, so remember it, right? tweet happened on it was the 21st? I remember last week, right? And then on later on Wednesday, you have like an initial patch from PHP saying this should fix it. And APT reposed. did not have any sick urgent security PHP embedded. No, you have to literally go to PHP and install it manually in a in a in a different fashion, right? so if your patching process so little relies on updating your security repos by that will do nothing. You need to know that you're running PHP and go and manually update a security thing. Yeah, it's it's it's not that trivial. I think one of the difficult things about PHP to shell as well, is that they backported the patch, right? So even if you are, you know, tracking all of your assets, let's say you're a really mature org and you're tracking all of your assets down to the version, like major, minor, and patch release versions, like this didn't update anything. So I think it's like 8.2.7, right, was where they fixed it. Mm-hmm. But it didn't update the version major, minor, or or release. So it's a a spreadsheet or an asset tracker with, yeah, I've got 8.2.7. No idea if it's patched. that's that's awesome. Right, right. And it's not like you can just like you in in some cases, like you can't just tell somebody, okay, well, you know, w it get off 8.2.7, get some other newer version, right? Because there's again, it's a platform dependency, right? If you're running an application that it's just simply not gonna work with anything but eight dot two dot seven, what do you tell someone in that case, right? And I just I think that's part of what was so ugly about PHP two shell, how it got leaked and and everything. Like this is something that has to be like It's very nuanced and detailed and the way that they go about fixing that without again creating other issues like notes to show. It like requires testing. And so like when something like this kind of leaks early and you're forcing them now, I believe they had to push out the patch Wednesday. They weren't initially planning on that. So that they bumped ahead like a full twenty four hours because of all the craziness. Like you're doing everyone a disservice. in the entire community as well as these other companies that are running this by forcing everyone's hand basically and use the second that tweet came out or the second anybody even released any information publicly, started that timer, right? So now it's like you you you're taking away the options for a lot of people. Yeah. Maybe this is just my way of throwing shade, but I think it plays I think it plays a little in the scoping question, especially on how much confusion it generates, but I just don't care about the like naming conventions that we're settling
now. Like the PHP to shell it I mean in an in a paper it does mention that you essentially go from a request to a shell, but it's not really happening ever. So so no it's just fear mongery at times, right? I mean it is critical. But the likelihood of exploitability is just so bizarre. And for this node as well, we just mentioned it. It's it's node because it ships a copy of it, but it's not particularly just node. So if you're just inventorying for node, then you're gonna have tons of blind spots on your attack surface. And I feel like this just proves that nowadays you just need a lot of inherent understanding of you know what's happening, like actually reading. Everything goes through it because seems like the namings they they don't they don't longer provide like any sort of insight. It's just more of a marketing sort of thing. So Yeah. Yeah. No, I don't know who to blame. I I mean the first one I can think it was is the log for J, the what was it, log for shell. I I'm sure there's one before that, but that's the one that always comes to mind. But like that one was like that it it it was like a mostly a command inject bug, right? Like like I know like I had written an exploit for it for VMware, or not the like ESX I'd like like login page. and like that one did. That one that one did give you a shell, right? So like that one was legitimate to to name it that way. But it seems like every it's like every bug now is just kind of chasing that same nomenclature of just we wanna be the next, you know, the very best that no one ever was. Yeah, yeah. I remember like we we should do we we we later should do top ten best names of the winning related, right? Like I remember Hartley. the the SSL Poodle was another one, or the TLS cipher if I remember correctly. It's like those were really cool names. Now everything is to shell, right? See, like I'm old, I'm like, you know, the the name that pops in my head every time and I'm thinking like order the the the most, you know, well known exploits is always like no MSO what was it? MSO eight six seven. Like that's that didn't have like a fun name. Like we just that's just what we called it. Yeah. Yeah. w for for that particular one, I remember calling it by the malware name, configure. sure. Yeah. Yeah. Yeah. That's that's true. That's true. Yeah, that's that's a good idea. Yeah. That's funny too to think about. It's like everything used to be everything used to be malware names, right? Everything all used to be like what was the worm or what was the what was the malware that came along with the exploit. The exploits weren't what we're getting named. It was like the campaign. It was the name of the like Trojan or the executable. And like people were out there spouting out like nonsensical lead code names and you're like whoa crazy yeah and and right now what we are tracking right and I think this is interesting in terms of this will be a great segue for the or our next story by the way but we we we know in the
past it wasn't past right it was like what late twenty two thousands right like early two thousands and we used to track by malware names then we started to get more sophisticated Sophisticated in the sense of how the research was done, vulnerabilities, and then we started naming the vulnerabilities as well. So we had the malwares, we have the vulnerabilities. And then we have the actors, which is like, okay, there is very dangerous people running this in organized cells, right? And and we are not tracking three names. We are tracking the actor name, we are tracking the vulnerability name, and we're we're still tracking certain malware names, right? Like particularly for ransomware. and I think that's that's the segue, right? Like. Let's talk about our three story completely. Let's let's move away from vulnerabilities and let's talk about memory, which is a new ransom group group. We don't normally talk about ransomware groups because you know, it's it's it's complex, right? It's complex to talk about them to talk how do they work. but what what I want to talk about them is one, the type of techniques we are seeing them using. there is one particular playbook they are running, which is running a running certain vulnerabilities, particularly CBE 2026 59310, which is kind of the major vulnerability they're running, attacking directly VMware v Center, attacking directly to the VMware V Center backup and system backup systems, right? what they are doing is trying to infect and retrieve a lot of information. What is interesting about this actor is, and you don't normally see this. Is the names they are throwing. So the one of the first names they throw was Open Text. Open Text is a big Canadian company. And there is UK industrial software maker, Aviva. both are quite big respectively, right? Maybe they're not like Sony. I know there is ransomware groups calling like Sony and what not okay. But they're big, right? They're they're important in that sense. Yeah. something that is interesting is how do you deal, or does it matter who is running Whenever there is an active campaign, one doesn't matter who is running the campaign, right? Like, is this a new actor? Knowing that it's a kind of not a known actor, can it's kind of new. Does that take relevance? Do you pay more attention? Less attention? Does that change how do you feel about this particular one? Two, the vulnerabilities they are using, right? Like and this is perhaps more towards Dylan. But how do you deal with vulnerabilities that are not necessarily new? But we see an explosion of how they are using it more today, right? This may this is a volunteer from earlier this year, but it's it's kind of ramping up on how they are using this win. Actors are their users. It's coming from the actors, which is kind of weird, really. Does it make it that kind of important? Is that a new emergen is that an emerging threat in in that sense? Well kind of yes and no, right? That's kind of the question. and three, I know Customers normally ask us, right? hey, we are seeing this activity from this actor. What do you tell them, right? does it matter this is a new actor? Does it matter what we know about the actor? Or is more like, hey, just one patch and that's it, right? Like what what should be the the the crucifaction? I think Yeah, I think yeah, like I think the the like the patch question is always just yes patch, always patch, right? But like but But I think it's like I I I've I've I've said before that like it it's it's useful from like like a CISO standpoint to know like is this actor how much does this actor care about me as a target of opportunity versus like a like they have a a passion for this industry for whatever reason, right? you know, if I'm if if if I'm in charge of the network for a hospital. I'm not gonna care as or I I rather I don't think I'd care as much for somebody that's targeting DOD installations or something like that. Right. you know, I'm still gonna be, you know, do my due diligence and and you know, administer my network and and do more of my cybersecurity things, but I'm gonna be like as like up to date and like and read in on the details of all of the other threat actors because at best I might be like an incidental target when they're just like scanning the internet for for somebody. yeah, Dylan, you you started to say something. I didn't mean to cut you off. No, that's all right. I've I've I think you stated it beautifully. Like I I think the only time like a real threat actor name really like matters whether they're a new player or not, is like credibility. But even then, you know, because we we've seen it before, like threat actors pop out and they're like I don't know, they're just selling previously breached data and claiming it as new and you're like, this isn't actually this isn't actually you guys. But other than that, it doesn't really like who they are as a group doesn't really impact anything that I'm trying to do from a specific vulnerability and playbook analysis and understanding, right? a name's just a name tag. your group could change its name a thousand times. What I care about is what you guys are doing. how you're doing it and you know, how you're pursuing or choosing targets. and I think that is what's really important. Like, yes, the names are I'm gonna catch heat, they're fun and like 50, but what really matters is the actions, right? And it's it's what you guys are doing. Yeah. On you know, like John said, it's it's really about understanding the threats within your scope that would want to come after you. I don't know that we I think we saw something previously about the like casinos getting hit last year. Like that was a big one. And they're like, it's like it was like basically the ocean's eleven of like online casino heist that you know. but it's like if you're a healthcare provider, it's like, yeah, those guys don't care about it. But there are others out there that do. And like you
should understand it becomes just a Just as important to understand their playbooks and TTPs because you better believe when something doesn't work, those groups are logging your response as well. Exactly. And they're like, this didn't work. It won't work on maybe this industry or this specific target. Try this other thing. And so it's just as important for you to be proactive about categorizing and logging the actions that they take so that you can understand. identify and respond to that threat appropriately. Right. Right. So I'm really happy that you guys are describing it the way I see it because I I didn't want to get my my spoon on it because I I'm kind of biased on how yeah I normally run these kind of things. But I think you you frame it perfectly. Right. Like it shouldn't matter like I have a problem with these thread feeds that sell you like a the the planet A spinning and real attack map is like, sure, it it it's it looks good, right? Like it's fancy. I like it. yeah. But Dylan and I are very familiar with the the what was what I can't remember the name of the map, but the map that's just like these are all the different attacks that are going on, you know, every single second of every day and there's just it's just yeah nonsense up It looks like a flight chart and you're just like wow. This is this is nothing. Thank you. Yeah. Yeah. It's it's yeah. I mean it's I'm sure that information it's it's useful to sell a product and that's about it. but like it it it it brings me to like it reminds me that of like the why probably the worst advice I think I've heard somebody give a SISO was they were concerned. They were concerned because they were getting a bunch of scans, right? They were like, hey, we keep getting scanned for this Vaughn. And it's it was like the same, I think it was like the same like couple of CVs. They were all for like the same product. They were just should we care? And they like, no, it's just scans. Don't worry about it. Right. and like context matters, right? That's it's like, yes, there's internet scanners that are just scanning everybody for everything all the time. And like, you don't need to care about those, right? Like you you everybody's getting scanned on my home network is constantly getting scanned for whatever, right? But for whatever agency, right? Yeah, yeah, exactly. From who knows, yeah, right. but like if again, like if like using like the the the hospital example, right? If you work in a hospital and like you're s for some reason getting s like specific targeting scans for like somebody's looking for like an exposed like smart printer or something like that, like and it's all this it's like it's it's heavily targeting to towards you, like That's that's a tipping point. that to me, that's you you have just been notified by an attacker that you are a target of interest. And that's information that you that arms you as a defender, right? that should immediately give you a new like or should immediately
start posturing you for like, okay, something's coming. Like either they cause like just because they're scanning you for for CVEs doesn't mean that's all that's in their playbook, right? If I like put on my my my bad guy hat, like if I'm going after one specific target of interest, of course I'm gonna scan them for the low-hanging fruits, the stuff that's already out there in open source. Like I'm gonna I'm gonna burn away all of my CVE trails and hold on to this zero day for just in case I need it. But you're not gonna see the scan for that, right? Like if if if you're just looking out for like the things that are big in the news, like you're missing so much of the picture when it comes to what what a an an attacker and I and I I don't even want to use the term nation state attacker anymore because that's not true. Right. Nation state attackers was like an old old world mindset of the only people that can launch these attacks are people with the funding and the resources and all that stuff. And like that was true maybe up until even a year ago. but AI has changed the game. Like any we've seen however many cases now of just random person with fifty dollar credit card and and an AI subscription and and they're launching attacks against large, large organizations and and succeeding. Like not just like launching attacks, like being very successful in their campaigns. Umapbox that one, yeah. I I saw this researching company the other day, they are starting to name some of them AIPTs instead of APTs, which is it makes a lot of sense, right? Like It it does it, but I almost I almost feel like that devalues the problem of Sure. Right. Like like I don't ever want to be like this is just an AI problem. Right. Like I I don't want I don't want any any threat actor to only be looked at seriously because now they're using AI. Like because guess what? Like the rest of them probably are too. Like you know what I mean? Shiny Hunters is probably using AI, even though I would not consider them like an AI PT. Right. Right. So something that is also interesting is And you mentioned it, John. The thing with and that's why it's always hard, what should I be looking like? The actor or the vulnerability, right? Because normally actors don't use for for both shell vulnerabilities that we just discussed, there is not a lot of report. There is certain cases where these companies like these companies I I mean like researching cybersecurity companies and firms, they are saying like, yes, we are seeing attacks on the wild, right? But there is no like named actors. And the reason for that is because normally these actors don't necessarily rely only on those fresh new vulnerabilities. They are using zero days. They are using vulnerabilities that no one knows about. They are using very old vulnerabilities. Why? Because you forgot to patch your broadly embedded server from many years ago. Right. Like it it's that that's why you don't normally see that. And that's why it's important to still mention it, to understand the TTPs of the actors and understand like. Are they attacking the he the health your the hospitals with PHP to show? Yes, okay, maybe you should no care about this actor, right? Otherwise, not necessarily ignore it, right? Because you it it's the same thing with vulnerabilities. You can't ignore the vulnerabilities. They will still pop up. You you keep patching. I think it's just as important to track you know, there's a difference between like tracking a threat actor that's within your area of operations and all they do is call in, right? And like social engineering is their big thing. Whereas like if you're actively being scanned, it's really important to review the logs. Like I I love when a client reaches out and they're like, hey, we're seeing this. And it's like, my God, someone's looking at the logs. Thank you. They're not just like, they're not just like, don't think anyone got in today. but they're actively reviewing stuff that flags and alerts. And it's awesome, man. Like you, but you have to take it a step further. You have to start building that knowledge base and actually getting contextual evidence of what that person's trying to do ultimately. Yeah. Because sure, let's say they don't get in now, right? But what are they looking for? Is it a specific technology? And like it's important to categorize that traffic, understand when it happened, like how frequently it's happening, and if there's a specific protocol or technology that they're targeting, because you need to give advice internally to like your product team and the CISO so that they can make informed business decisions when they go to maybe change tech stack later or rotate off a specific vendor. And, you know, if you're not keeping That and I'm not saying you have to have crazy historical notes worth billions of dollars of storage. It's just like logging, understanding the frequency, like keeping metrics and understanding that. And like it it's gonna pay dividends in the long run. Right. It's not the cool sexy stuff, man, but it's like safe. And that is cool and sexy. And it's a yeah, right. Like I I really not losing a bunch of money. And I'm sure businesses feel the same way, right? And like and like to your point, like cyberspace is like one layer of the stack here we're talking about when it comes to an attack, right? You know you mentioned like, you know, the scan logs and stuff like that. That that's not taking into account like, you know, the the meat space aspect of like are they doing any sort of social engineering? Right? Are they like like again if I am specifically trying to hack into this company, this organization, whatever, there's a multifacet approach here. And it's not gonna be just me sitting in my chair across, you know, however many miles and just throwing packets at your edge devices. Like if I'm trying if I'm trying to hack in, like that's that might be the low-hanging fruit aspect of it, or not. That
could be what I fall back to. I may have already established some other mechanism. I may have already social engineered my way in and I'm and I'm just looking for another back door in case I get closed out on this side. Right. there's so many of those factors that I've been there for. Right. Yeah. Exactly. Like it's and and it and it's funny because it's like as an attacker, like putting on the attacker brain, like It the the the things that people or the sorry, the things that defenders catch and and and they they scrub it and they're just like, Yeah, cool, we we got rid of this actor, we got rid of of this threat. we have this like human bias to pat ourselves on the back and just be like, We got it. We got rid of them, right? We close we we patched this vulnerability, you know, yeah, we found, you know, we we did our our incident triage and we and we we found the we found the way in. not realizing that what else is there, right? Yeah. No, that's that's a perfect way to frame it. and I think this is this is a great opportunity to bring the the all all of the topics together, right? there is two very catastrophe vulnerabilities, right? not really, jokes aside, two very bad vulnerabilities. There there is a new actor and dozens of them, right? On on the queue. There is a lot of things happening. This is a spooky season, spooky vulnerabilities, spooky attacks. So what are and and I hope to get a a response from each one of you like if a CISO asks you what what changed this week and what they should do on Monday, right? Like new new week, Monday, what they should crab out. What's your two sentence answer to them? I'm not going first. Don't put me on the record like that. I I I will get it started. My my recommendation is make sure you are up to date and that your strategy involves doing an inventory of your assets, including assets at the hardware and software level, and make sure to keep up with patches and updates. And after that, you can start monitoring, right? What will get a solid baseline, a sort a solid, not even severe security posture. An IT posture, right? a solid IT posture before you move forward to the next step. Make sure you have that and you will be on a better place. I will not say you will be safe, but you will be on a better place. Yeah, you'll be safer, right? It's all just those are good. we're trying to right there. That's not no further. No. That's me. Good at all. Cov covers a lot of it, right? I think number one is being be proactive, not reactive, right? And and that encompasses patching, right? And and and you know, policy making and stuff like that. you know, but I think it's if if you have, you know, understanding that like not every business is gonna have the spare cycles to do stuff like, you know, source code review, but that's something that if you if you know that you're like what your tech stack looks like, if you've done all of that aspect of like, I know I'm running these things, that's maybe something that you
should leverage as part of your defensive posture is like, you know, getting some sort of a having some sort of source code review for your own product or for a product that you use. and then Hmm. I guess the second one would just be like s I mean like stay as as up to date as you can on what's happening and and and what's happening specific to your industry. you know, don't it th there's there's too much information, right? Like that we we we it's a good problem to have, but there is too much information and it's easy to get overwhelmed by like we just had last week with just a wave of of vulnerabilities. But as critical as both of those, both of those families of vulnerabilities were, you may not be affected because you're just not like like even if you're employing a pro a product that uses PHP and uses, you know, node, if if those are not exposed in the right way or if your configuration is different, like you may not need to care. And that's an important thing to know too, is like, yeah, you need to update, yeah, you need to do all these things. But knowing How those exploits are reachable, like the actual true, like true to the term exploitability, I think is an important thing for for IT and defenders and Civos. Mm-hmm. Yeah, I think that'll be no Emmy go next because I'm gonna let I d Dylan's gonna have a good answer. I want Emmy to go. no. So the way Dylan doesn't steal Emmys, that's what I'm asking. Well, to avoid saying the same thing, for this particular case, I would just say that. prioritizing PHP to shell, just giving not only because of the vulnerability itself, but just the sort of reception that he got worldwide. just even if you're not exposed, if you have any sort of service, then your clients are gonna reach out and asking all of these questions. Hey, are we affected? Hey, is this an issue for us? And inventorying and cataloging is quite complex in this in this one. So definitely a priority Just to make sense of the whole thing. And I think that this is this has been our new slogan, which is like sifting through the noise. Like, yes, you should always be monitoring for different advisories and everything that's dropping. But at the same time, asking the question of how do we go from this advisory to actual exploitability? Because that definitely changes what you should be prioritizing and what you should be on the lookout. Right. John and L me really touched on context a lot matters, right? it's easy to get caught up in a whirlwind of hype. again, doing the basics really helps, right? So like that asset inventory, but not just as much that like physical asset inventory, but what's actually running, like that software bill of materials on those assets. and again, knowing knowing the information that like Others can't extrapolate for you, right? Understanding that there is a difference between what can be seen externally and internally is is a huge leg up and and gives you an advantage as defenders, right? Being able to see what you have internally, whereas others might just be blindly throwing stuff and give you a tip off that, like, hey, this wave of crap is coming. but understanding the context too, like John stated, More and more, I think it's gonna matter a lot of the CVEs and ETs, that we're publishing now. There's so many different niche like preconditions. And it's we're no longer in the day and age of if you're running this version, like you're definitely vulnerable. It's like you and I I think part of it has to do with, you know, how we're just releasing CBE left and right for any old claim because we can't keep up. But again, it's like some of these aren't proven. Some of it's just noise. and and understanding like really looking at the things that matter for your attack surface and understanding the preconditions that have to exist for exploitation. Cause there is a clear difference between vulnerability and exploitable. And and it matters. Right. And that's precious time that you could be saving or dedicating towards patching something else. Yep. Hundred percent. Agreed. I literally agree with everything that you you everything that all of you said. that's that's great. I think the CISO will be happy, in my opinion. I think that's a good place to wrap up this MSS edition of initial access. the message From the attackers is simple. Patch me if you can. Okay, that that was a better joke, right? Yeah, I like that better. remember the the day zero can come from this closure from eight months ago from a tweet, right? yeah, it's the spooky season is not over. I hope we don't get anything else, but I I can't tell. Yeah, we we won we all we we'll get hope. again, I was joined today by Jonan's Emilio Gallegos and Dylan Sparks. My name is Sergio Villegas and yeah, thank you for watching. Remember, find us in you can find us in the Discord, Picture Fox, and the subreddit as well, in case you want to engage with us. And yeah, let us know if you like these episodes or what you want us to be talking next time.