Video: 2024 Annual Threat Landscape | Duration: 3660s | Summary: 2024 Annual Threat Landscape | Chapters: Welcome and Introduction (0s), Team Introductions (214.315s), Cyber Threat Landscape (338.53s), Threat Actor Evolution (1463.69s), Russian Sabotage Operations (1619.695s), Enforcement Impact Analysis (1923.155s), Positive Security Modeling (2149.96s), Persistent Influence Operations (2394.695s), Emerging Cybersecurity Threats (2610.05s), Shifting Threat Landscapes (3034.895s), Effective Security Practices (3255.815s), MFA and Detection (3315.855s), Taiwan Cable Incident (3440.92s), Blackbosta Logs Revealed (3506.775s), Closing Remarks (3617.865s)
Transcript for "2024 Annual Threat Landscape": In here. Alright. Hello, everybody. Welcome to Reporting Futures webinar discussing the 2024 annual threat landscape. We're we're super excited to have you joining us here today. As we have folks coming in and settling in, we will be pulling up a a Slido so that folks can be interacting with us, as we have folks that that are still joining. So as you're coming in, please do feel free to scan the QR code, join us at at slido.com, and answer the question that you see post here on the screen. So based on, how your 2024 went, you know, what were some of the biggest threats that pose the greatest concern to your specific organization throughout 2024. Once we have folks that have come in and and have scanned that QR code and and answered some of the questions, we'll start to see a word cloud populate based on those responses. So excited to see those as as y'all join us and and join the conversation there. A little bit of housekeeping before we go ahead and get started. I will note that this webinar is being recorded. All registrants will receive a copy of the presentation that will be shared via email after the live session has concluded. If you have any questions for any of our speakers or our panelists, please feel free to put them into the Q and A tab that you should see on the right hand side of your screen. We will also be using the documents tab that you should see on the left hand side of your screen, to share some additional resources that we'll be referencing throughout the presentation. So thanks so much. It's great to see these answers come in. It seems like ransomware phishing pose some of the greatest greatest concern and and the greatest threat to various organizations throughout throughout 2024. Alright. Okay. A little bit of of, again, some housekeeping as as we have folks getting started. Welcome to Recorded Futures webinar as we'll be discussing the 2024 annual threat landscape. I have a few of my esteemed colleagues that I'll ask to to join the conversation and join me here on stage to run through some quick introductions. So, I'll I'll run through a little bit of run a show, and then we'll run through the intros for the folks that we have on the line to speak to you about the 2024 annual threat landscape report. We'll start the conversation with, again, a quick round of introductions, and then I will pass it over to my esteemed colleague, Megan Keeling, who will be running through kind of an overview of the 2024 annual threat landscape report. After Megan run through runs through those slides, spending about fifteen minutes on on that part of the discussion, we will then switch to a panel discussion for about twenty, twenty five minutes where I'll be posing questions to the folks that you see here on the screen about the what we saw, throughout 2024 and and what we anticipate for for 2025 as well. After the panel discussion, we will then move into the q and a part of of the conversation to to close the hour. And again, please use the the q and a tab that you should see on the right hand side of your screen to input some q and a, and and any questions that you might have for for the panelists that we have with us together. Alright. Kicking us off with introductions. My name is Maggie Coleman. I'm a senior manager at Recorded Future, and I lead a team of analysts who put together various types of intelligence reports, including the annual threat landscape that they will be discussing here today. My team also puts together various intelligence products and reports that we deliver directly to recorded future customers based on their intelligence needs. And, Megan, I'll pass it over to you for a quick intro. Great. So, I'm Megan. I am a senior strategic threat intelligence analyst with the InSight group, and I work closely with Maggie on coordinating, putting together this report from recorded future data as well as InSIG threat intelligence. My intelligence specialty is organizational risk impacts derived from cyber threat intelligence. Amazing. Thank you so much, Megan. Alex, over to you. Hey, everyone. My name is Alexander Leslie. I am a threat intelligence analyst with Insig Groups advanced cybercrime and engagements team, and my team is focused on all forms of cyber dependent and cyber enabled crime. Awesome. Thank you, Alex. Matt, over to you. Thanks, Maggie. Hi, everyone. My name is Matt Mooney. I'm the director of the global issues team within Recorded Futures Insight Group. Our team focuses on strategic geopolitics, threats to physical security, and influence operations, and we produce a range of geopolitical intelligence products, both short and long form. Thanks, Meg. Phenomenal. Thanks, Matt. And Derek, over to you. Hi, everybody. I'm Derek. I'm a senior threat intelligence analyst with Recorded Future. I've been here for about a year, and I have been on Maggie's team, specializing in, threat landscapes, strategic threat landscapes for our Recorded Future customers across pretty much all, all industry verticals. Awesome. Phenomenal. Thank you so much, Derek. Alright. And, Megan, I'll pass it over to you for an overview of, what we observed occur across the cyber threat landscape throughout 2024 and also, a little bit speak to what we can anticipate for for 2025 based on the findings that we've included in the report. So, Megan, over to you. Great. Thank you, Maggie. So as Maggie said, I'm going to introduce our six key findings that were included in our cyber threat landscape for 2024. I'm gonna be going through these relatively quickly because I wanna make sure that we have plenty of time, for our panel to discuss these and other topics that come in through the q and a. So starting with our first trend, the cost of credential theft is higher than ever, and this is driven by two kind of interrelated trends. The first being a 40% increase in software as a service application adoption, throughout industries since 2021. And not only have the number of SaaS applications increased, but the function that those applications are doing are increasingly important, increasingly integrated into business critical functions. This is feeding into an increase in credentials, obviously, to log in to some of these SaaS applications. And if we look over at our threat data looking at nation and credentials exfiltrated using information stealing malware, more affectionately known as info stealers, you can see that there's been a 24% increase in credentials stolen per device, year over year. And I wanna take a quick opportunity to jump over to a example of what this looks like in our identity, intelligence platform. So we look at, to get that number and to get other information on this for the annual report, we sorry. Just pulling this up really quickly so I can show you. We look at compile a bunch of information logs, which, obviously, this one is anonymized because these all include personal information. But we could see not only the credential that is stolen, but also where it was stolen from and, in particular, some of the device information too, which and the malware family that is responsible for stealing it. And when you put all these together, you could start to get some of the trends that help inform, you know, where these, how these malwares are operating. And in addition to seeing that increase in credentials, we also note that those credentials are coming from personal devices, not just enterprise devices. So they're outside of most enterprises' ability to kind of keep an eye on this, but they still have an impact on enterprise SaaS applications. One, most people do some form of business on their personal devices, but also two, some most credentials are often reused across, different applications as well. So if you steal a credential from, a personal device, it may be able to be used for an enterprise application. And we saw this occur during the Snowflake breach that occurred last spring. Snowflake is a, a cloud file sharing application. And what happened was a 65 organizations saw their Snowflake cloud instance breached, as a result of the threat actor compiling data from information stealing logs like the one that we just took a look at, and using those to break into their cloud data storage, exfiltrating really critical information. This was enabled by the lack of multi factor authentication, which more easily allowed the threat actor to take advantage of these exposed credentials. However, the threat actor also had malware that was specific to the Snowflake instance. So once they were able to gain that initial access from there, they were able to pivot and get after some of that more high profile data. In addition to the lack of multi factor authentication, misconfigured single sign on also plays a role in maximizing how threat actors are able to abuse valid credentials for SaaS applications. You could see that, in a lot of cases, organizations will try to combine all of their SaaS onto a single sign on. However, sometimes users are accessing SaaS applications that are outside of that, whether it is a misconfiguration error or just something that attack surface. So moving into another, our second trend, we look at a series of MITRE tech tactics, techniques, and procedures across all observed, or all of the observed cyber activity, and saw the greatest increase in both defense evasion and discovery. And these are two attack attack tactics that allow threat actors to essentially hide in plain sight. And discovery in particular is valuable for this because it allows threat actors to see what processes are already running, what programs an enterprise is already using, and then take that into planning their lateral movement and other forms of, exploitation. One potent tactic deep dive that we especially saw was exploiting remote management tools observing a 214% increase in incidents. These were enabled by direct vulnerability exploitation such as the ScreenConnect vulnerability that allowed the threat actors to take control over this. However, we also saw social engineering as well, in particular using the help desk scam to get victims to intentionally download either a, malicious version of a legitimate tool, so kind of using that tool for malicious purposes, or a spoofed tool, that was malware disguised as a legitimate tool. And, again, if you're able to disguise your, help desk scam using something that is already in the environment, it makes it far less likely that defenders are going to be able to pick that up. So these tactics were used by both nation state actors as well as cyber criminals. But for the next few slides, we're going to focus into specifically what cyber criminals were doing to maximize illicit profits. So 2024 was a really big year for law enforcement operations against, criminals. So, some good news to kick off this slide before we get into the less good news. In particular, we saw an operation in February, Operation Kronos, which was, a coalition of law enforcement operators from The United States, United Kingdom, and others who took down LockBit, who was an extremely prolific, ransomware affiliate ransomware as a service affiliate group. Not only did they account for 34% of all attacks in the first half of the year, but they accounted for, I think, 22% in the previous year and were also the top actor the year before. So definitely a major space on the ransomware scene. However, once LockBit or following the takedown of LockBit, which was involved the arrest of two of their leaders as well as taking down their infrastructure, we saw the number of variants significantly increase in the second half of the year. So we saw 72 new variants, which are new ransomware families that had not previously been observed, in h two. And this is especially important when you consider that there are only 34 or 32 new variants for 2023 total. So definitely a big jump there. And these new variants were extremely active. We saw about the same number of ransomware attacks, being carried out in 2024 as in 2023. However and this isn't in the report because this came out both from us as well as, cybersecurity analysis group chain analysis, just last month. But it's important to mention that the actual proceeds from ransomware extortion operations has actually decreased. So it decreased by about, I think it's 800,000,000. So suggesting that these new groups are actually less effective than, some of the more organized affiliates prior to the takedown. That being said, big game hunting is still alive and well. We saw a $75,000,000 ransom payment, which was a record setting payment in July of twenty twenty four, that was paid out to a group called Dark Angels, by a Fortune 500 company. So that goes to show if you're able to make the right exploit, if you're able to attack the right spot, there's still, ransomware can still be a lucrative form of cybercrime. Looking at some industry trends, we can see how threat actors attempt to maximize their return on infection by going after companies that they think will be most likely to pay out the extortion. And we could see, manufacturing is the top industry in extortion claims for this is actually the third year running now that they are, claiming this dubious title. This is because rent or manufacturing has a particularly low tolerance for downtime. It's very expensive to take a factory offline with some estimates anticipating around 33,000 per minute for certain types of factories. So, really large impact there. There's also an inconsistent network segmentation that tend between information technology and operational technology that gives threat actors the opportunity to, if not directly impact this high highly consequential operational technology, at least to threaten to do so, by getting into information technology, and enterprises are not always aware of how those two connect. So, these two factors give threat actors a greater opportunity to try to demand these really high payouts. In other forms of cybercrime, we can look at information from our cybercriminal forums. These are postings on, dark web networks, which offer initial access, which can be used for data exfiltration as well as extortion, as well as database sales. And we've seen database sales also significantly increase in 2024. This could be for a number of reasons. One potential contributing factor is that stolen data is extremely useful for committing fraud. Fraud has been a multibillion dollar business and is increasingly on the rise. And depending on the type of data, that is found in these databases, these could be especially effective, for carrying out fraud fraudulent claims. One high priority database that or set of databases in telecommunications, for example, which, tends to run pretty or tends to be one of the more highly priced, so therefore, potentially meaning it's more valuable. But these types of data could be used for SIM swapping attacks as well as for bypassing multifactor authentication. K. So moving away from the criminals, I wanna talk about a couple of trends that we've seen in nation state activity, before we move on to the panel. So we've seen threat actors targeting critical infrastructure in, relatively new and unprecedented ways, and these are driven directly by geopolitical conflict in between nations. In February, the United States government announced that Chinese state sponsored group, Volt Typhoon, who is, related to, like, very likely the Chinese military, has been observed pre positioning in US critical infrastructure networks. And, the US government distinguished this between espionage, which is obviously focused on the exfiltration of critical information, based on where this activity was observed. They assessed that this was not to exfiltrate information. It was to disrupt, electricity communication or other essential utilities, on in The United States. This could be to demonstrate the capability to do so or as an actual prepositioning in light of escalating conflict. We can discuss this more in the panel, but The United States and China have had a variety of issues over which they have clashed, but most notably around China's territorial ambitions in the South China Sea as well as in Taiwan. And it's also notable that we have seen significant threat actor activity in Taiwan's critical infrastructure as well. Pivoting over to Europe, we could also see similar patterns, particularly in Ukraine where we see very direct Russian linked attacks on critical infrastructure causing disruptive actions, but notably in NATO aligned Europe, so who's distinctly not at war with Russia, and protected by article five protections, we see much more indirect targeting of their critical infrastructure networks. We've seen, hacktivists who generally align with pro Russia views go carry out relatively low sophistication attacks against critical infrastructure, throughout, throughout Europe, particularly targeting h d HMI devices in, water treatment facilities and other, operational technology. We've also seen physical attacks be carried out as well against different critical infrastructure entities. So carrying out, sabotage, arson, anything to cause a kind of physical impact, as well as notably activity in the Baltic Sea attempting to disrupt undersea cables, again, with that veneer of deniability, using ships with, multinational flags as well as attempting to, you know, make it seem acts like an accidental operation, all to kind of carry that layer of deniability while still causing disruption in critical infrastructure. So the last set of nation state activities that we'll discuss are malign influence operations. So, a different kind of online operation that is distinctly scaled up by the increasing adoption of generative AI into some of their tactics and techniques. And we've seen generative AI be used across, not only different, countries but also in multiple operations. For example, Russians or the Russian linked operation overload has been observed using both deepfakes of known and fake personas to attempt to spread pro Russian narratives around, US and French election as well as, the the Olympic games in Paris Last summer. And notably, they use fake personas that are acting as whistleblowers, to attempt to overload researchers and, journalists with these, potentially issues to investigate, thus muddying the waters between, you know, what is a real investigation and what is, and what has just been made up. Similarly, we've observed Iranian influence operators attempting to stoke cultural divisions in Israel by using not only fake personas, but as well as coordinated and authentic engagement to attempt to increase cultural divisions, not only by starting fights kind of between these, fake figures that are operated by the influence operators, but attempting to bring in, real Israeli citizens as well into these, into these conflicts. And then finally, we've observed Chinese influence operations, again, using generative AI to scale content, particularly in using multiple languages to spread, both content that is against, China's adversaries, in particular, redirecting hashtags associated with 10 Tiananmen Square toward content that is critical of police violence. We've also seen a lot of pro China content as well-being spread through these, inauthentic channels. And that is our very quick overview. So I will turn it back over to Maggie and excited to see the rest of my colleagues out here to answer your questions. Amazing. Yeah. Thank you, Megan, so much for for running through those those high level trends. I'm excited to pivot the conversation to a bit of a panel discussion where I'll be posing some some questions, diving in a little bit deeper into some of the topic areas that Meaghan ran through. So, Meaghan, thank you. Thank you so much. I wanted to start with a kind of high level question, specifically asking some of the the experts that we have on the phone, what was the biggest shift or change that you saw in threat actor behavior over the last year throughout 2024 based on your your area of expertise? And and, Derek, I'd I'd like to pose the question to you first. Throughout 2024, what was the biggest change in threat actor behavior that the folks that we have on the line need to be aware of? Yeah. Thanks, Maggie. It's a great question. So from my perspective, one of the biggest shifts has been sort of the dedicated use of AI in the development of, specific tooling or, sort of the expedition of the kill chain for threat actors. Right? The beginning of the year, we kinda started off with, sort of transparency reports from OpenAI and similar companies saying, hey. Threat actor a has been using this to develop scripts, things like that. From industry reporting, we would see, hey. You know, we observed a bash script that was executed on the command line where it had all the kinda hallmarks of GenAI that anybody that's played around with ChatGPT would be able to recognize in terms of, you know, being syntactically perfect in terms of the spacing and indentation of the code, perfect, English language, accurate descriptions, each line as to what it does, things like that. But towards the end of q four, we saw kind of a a a bigger shift towards sort of larger payload development. Two of the kind of biggest that come to mind are the, Radamantha Steeler, where you saw the use of AI in the form of, image recognition, where they were able to sort of pull, like, cryptocurrency wallet related data, using AI overlays with, images. And then we also saw the emergence of the funk funcsec, ransomware group where their entire encryptor, is likely developed using AI. Right? So there are all these same hallmarks. You see kind of, a discrepancy between the encryptor itself and their interactions on forums in terms of their command of the English language, in terms of their, technical understanding of what their payload is doing. Right? And so that for me has been the biggest shift that I've seen is going from kind of light experimentation and, hey. I don't know how to perform this particular enumeration. Write me a bash script for this, all the way up through, integration into malicious payloads. Awesome. Yeah. So kind of just increasing use of of AI specifically and kind of more sophisticated use of it throughout the year. Awesome. Thank you so much, Jared, for for walking through of that. Now I I guess pivoting, Megan, back to you. From your perspective, what was the the biggest shift that we observed throughout 2024? Yeah. I think for me, this came from working on both the 2023 report and this year's report. We talk about SaaS applications in both of them. Again, that makes sense kind of given their increasing prevalence in a lot of enterprises. But last year was really focused on the exploitation of zero days, and we saw a big uptick in zero day exploitation in these tools that we didn't really see repeated this year. Instead, we kind of saw threat actors shift more toward those valid credentials using social engineering, as well as more, quite not using native processes and native tools and legitimate tools in order to carry out exploits. So I think for me, that is one of the bigger shifts is trying to remain undetected by doing things that you would likely that you're more likely, this kind of using the tools as they are intended to, except as a malicious actor. So in order to evade, different, cyber defenses. Awesome. Yeah. And then and I know that that kind of evasion of of defenses and and specifically, like, some of the EDR mechanisms is something that we dive dive into in the report in detail and and actually have a question on on that for later. So excited to get into that a a little bit more. Matt, what about from your perspective? What what did you see as as the biggest change in threat actor behavior throughout throughout 2024? Totally. Thanks, Maggie. So I really focus on Russia's escalating attack plotting against Europe and the West over the past year, exemplified by the assassination plan targeting the CEO of German arms manufacturer Rheinmetall as well as attempts to send incendiary devices on air cargo flights to The United States. So Moscow Moscow has very likely intensified its sabotage operations against Europe over the past year as part of its hybrid war warfare strategy. So in this, it's basically intended to augment its conventional capabilities and its operations in Ukraine with deniable and subversive actions against Ukraine's allies. The very likely expanded scope of Russian sabotage operations in Europe poses significant, threats to European critical infrastructure, military capabilities, and political stability. So, we published research late last year that outlined how Russia's recent sabotage, TTPs, or tactics, techniques, and procedures very likely remain consistent with Soviet predecessors, with the obvious exception of technological advances. So this research identified several plausible scenarios of Russian sabotage in a specific time frame, June 2024, which Megan alluded to in her presentation. So the methodology we used behind these assessments was based on multiple criteria, including the nature of the target, the use of rudimentary tools or tactics, and the likely impact on a country's military, economic, or political sectors. So Russia's sabotage operations almost certainly seek to destabilize NATO, degrade NATO's warfighting capacity, and disrupt NATO sources and support to Ukraine. More recently, Russia has been accused by multiple European governments of its involvement in the intentional severing of sabotage of sorry, of submarine cables in the Baltic Sea. And so for context, there are approximately, you know, 100 to 200 of these severings that occur annually, the vast majority of which are due to accidental human activities such as fishing, or natural events such as underwater rock slides or earthquakes. And while we don't have comprehensive stats detailing exactly where these all take place, we do know that there are 600 plus cables operating around the world today, and these total 1,400,000 kilometers in length. By comparison, the Baltic Sea is home to approximately three dozen cables, the longest of which is the Sea Lion one, which stretches nearly 1,175 kilometers. So incidents in the Baltic Sea are increased notably over toward the end of twenty twenty four. In November, the Chinese boat carrier Yipeng three dragged its anchor for approximately 100 kilometers after departing a Russian port, severing cables connecting Finland, Germany, Lithuania, and Sweden. And then on Christmas Day, the Eagle s oil tanker, which is suspected of being part of Russia's shadow fleet designed to abate sanctions, dragged its anchor for approximately a hundred kilometers after leaving the same Russian port, severing four fiber optic cables and one power cable. And we've seen this trend continue into 2025 with numerous cable severings in the past six weeks, including the Sea Lion one again several days ago. Sorry about that. We're still on mute. Thank you, Matt, for for that really comprehensive answer. Really, really appreciate it. Alex, I I'd love to to pivot to you and and get your perspective on what was the biggest shift that we observed in threat actor behavior throughout 2024. Yeah. So Megan already touched on this a little bit, but in terms of cybercrime shifts, the biggest shift that we've seen is that decentralization. So the pivot away from centralized forums, marketplaces, messaging platforms, the the relative decline of the as a service business model. So ransomware as a service, infrastructure as a service, malware as a service, especially relative to 2023, the fragmentation and the splintering of preexisting groups, especially major groups. We can talk more about Blackbosta later, which seems to be on everyone's mind, and the rise of independent operators. So smaller groups, affiliates that may have broken off from a larger ransomware group, and and, everything that comes from that. So there's a lot of factors driving this decentralization. We can talk all about law enforcement operations in a little bit. That's a topic in and of itself, but internal divisions, leaks, right, Blackbosta, and even Dockspin within the, last week. The pivot away from Telegram with the arrest of Pavel Durov, later, in in the later part of last year, and all of these shifting centers of gravity within the ecosystem. So I see a lot of pros and cons here. Two things that I I really wanna emphasize. The cons first being, you know, this is making it very difficult for researchers to monitor threat actor behavior, especially when when it comes to human intelligence collection. This decentralization is quite difficult as actors move to, private sources, special access sources, and so forth. But then the flip side of that being this decentralization causes a significant amount of operational disruption to independent actors. It's much higher operating costs. You don't longer have the the broader ransomware group to provide you with infrastructure, tooling, manuals, and so forth. You're kind of on your own. So it's a double edged sword for both researchers and actors, and and I'm sure we'll dive more into that later. Amazing. Yeah. Thank you so much. And and is it a great kind of tee up to to the next question that I wanted to ask, Alex, which is related to some of the knock on effects of the law enforcement operations that we we saw occur throughout 2024 and also into the beginning of twenty twenty five. So can you speak a little bit more to that? What are the long term impacts of some of the enforcement activity that we've seen, if any? Is this kind of a game of whack a mole or are there real positive changes that that we've seen come about as a result of some of this activity? Yes. So you asked about long term implications. I think first and foremost, I need to emphasize that the short term wins, the operational disruption that comes from short term operations is very clear. So those groups that are disrupted, say it's Alpha, LockBit, Phobos, and so forth, those groups immediately have a significant operational disruption that they have to become adaptable to and overcome, following a law enforcement action. Those wins compound over time, and they have ripple effects. So so to kind of emphasize the implications long term, the ripple effects of these law enforcement operations are reshaping the underground before our very eyes. We are seeing a whole host of new players. Megan mentioned that we had over, 70 variants in the second half of twenty twenty three. We are seeing over three dozen new leak sites in last year alone, rise of independent operators that don't use leak sites at all, some of which don't even use ransomware, and they just focus on data theft and data extortion. They don't announce their activities publicly. The center of gravity is shifting. So when we have LockBit being the biggest player in 2022 and 2023, that power gravity center is shifting over to ransom hub, and it's shifting over to other smaller splinter groups associated with these former major cartels. The second big effect is going to be long term deterrence. So the cadence of these operational disruptions, the cadence of these activities, especially as they increased in pace throughout 2023, has had had a significant deterrent factor on preexisting groups that were affected by disruption. So you see LockBit as an example with operation Kronos hemorrhaging affiliates, hemorrhaging the amount of victims, reposting victims in order to claim that they're they're still active. Infrastructure goes down, decryption keys released to victims. This significantly affects the group's ability to to, conduct attacks, and it deters them from future attacks. And then the last the the last big implication is that change in in group behavior, that pivot away from ransomware purely to data theft and data extortion. All of this is to say this this builds momentum. This is compounding over time. I kind of like the whack a mole analogy because it has some positive nuance to that in in in terms of those incremental wins are all small tactical wins in a very long war. Those incremental wins, that whack a mole builds morale within law enforcement agencies and builds collaboration. It retains expertise. It fights for funding. Yes, those geopolitical safe havens still exist. Russia is still harboring criminals. We it's it's it's, you know, remains to be seen how that relationship with The United States and US law enforcement evolves over time, but I need to emphasize that 2024 was all about progress and not victory. It's it's tilting in favor of the good guys. You know, we are we did the the first quarter just finished in week one. We got a long way to go before the Super Bowl. Right? Like, there is a lot of of activity that we have to to hammer down. So keep your eye on the ball, but overall, very optimistic about law enforcement action last year. Amazing. Yeah. And I love that reframing of whack a mole. I feel like oftentimes when we talk about that, it has a really negative connotation. So I I think that that's a great a great perspective and, definitely a positive as we look to the rest of the year. So thank you so much, Alex, for running through that. I know that as you were chatting, we had some really great questions come in. So excited to dig into a few more of those key themes that you mentioned when we get to the Q and A portion of of the conversation. Shifting gears a little bit, as Megan ran through, we did see an increase in some of the defense evasion, and bypass of endpoint detection and response mechanisms throughout 2024. Some of this activity was fueled in part by the increase of abuse of remote monitoring and management or RMM tools by threat actors, which we saw surge throughout 2024. So, Derek, I I wanted to pose this question to you. In light of some of these changes, what are some strategies that defenders can adopt to detect methods, to that are intended to bypass, you know, traditional defense mechanisms and and and ADR? Yeah. That's a it's a really great question. And so from from my perspective, sort of the bottom line on this is before you can really start determining what bad looks like, You really need to know and understand thoroughly what good looks like. Right? And this gets into your, kind of the the more direct answer, which is more positive security modeling. Right? What you want to start considering is that, as threat actors kind of move away from, executing code in places that we're looking into places that we're not, That's always gonna shift. They're always gonna shift once you sort of find the new place. So, really, the ultimate kind of, end state for this is kind of that exit, like, execution state modeling, right, where you understand what your assets are supposed to look like, what they're supposed to behave like, how they're supposed to be configured. And when you detect and this is, you know, for, you know, even for running processes as well. Right? And in that way, once you start detecting that a particular asset has been modified outside of its expected state, outside of its known good configuration, then at that point, you have, you know, a a an indicator of change. Right? At that point, you have your sort of, detection models take over and start moving down the path towards, you know, triage investigation, mitigation remediation, etcetera. Amazing. So kind of that that baseline establishing what, you know, what known good is so that you can then detect some of the the malicious behavior. Awesome. And that and that even, you know, goes all the way down to the individual user. Right? Because when you have, some of these kill chains where they're using, say, command line, PowerShell, that kind of a thing, that could be totally normal for a a developer or a threat analyst or something like that. And then on the other hand, if you have somebody in a different department that's just never used PowerShell before, never even opened that up, and now you start seeing bash scripts come across, that's where that deviation from baseline, matters. That's where the kind of money, gets made there. And that's where this sort of positive security modeling execution state mapping for your assets kinda needs to go. Awesome. Yeah. That makes a lot of sense. Thank you so much, Jared, for for walking through that. Pivoting a bit to influence operations, which Megan discussed in in her overview. Obviously, throughout 2024, there were various geopolitical events that that occurred that, facilitated some of the the, or I guess, provided ample opportunities for some of these influence actors to attempt to mold or sway public opinion by the way of of influence operations. So while our research showed generally limited engagement with many of of these influence operations, Matt, I I wanted to pose the question to you. You know, why in the face of of some of that limited engagement, do do you think that we see these influence operations persisting, expanding over time? Yep. Certainly. The short answer is, you know, because they're cost effective mechanisms for polluting the information space online, inflating perceived impact while offering the potential for breakout influence events, and more on that in in a little bit. At a more strategic level, influence operations are often layered with other threat vectors such as offensive cyber operations. So it's difficult to calculate their impact based purely on engagement alone. Our research over the past year, has detailed numerous influence operation campaigns by state actors. Megan alluded to some of these, most notably Russia, China, and Iran, targeting elections in France, United States, as well as the Paris Olympics. And so they do that in an effort to destabilize Western democracies, inflame political and social tensions, and sow discord among the general populace. Russia, in particular, has continued to double down on influence operation activities by leveraging generative artificial intelligence to scale, production output, create an overabundance of inauthentic content, and then automate its distribution through social media. Believability is very likely not always the primary objective in these cases. Overwhelming the information environment with junk content to de degrade the overall aquatic discourse is almost certainly an end in itself. Now these campaigns also seek to create the belief that they're more impactful than they actually are, just a tactic referred to as perception hacking. Now for one example, a campaign that we track is Operation Overload. This is a Russia aligned influence operation, consists of inauthentic news and fake fact checking content that impersonates legitimate news sources. The operation seeks to establish credibility by abusing the brands of trusted media organizations, using fake expert testimonials, and creating localized, allegedly eyewitness content that's difficult to verify. The threat actors behind these IO networks, they're nimble, and they've demonstrated strong capacity for tailoring their output to key geopolitical events and themes. So a very good case in point, unfortunately, was a viral video that was produced earlier this month targeting USAID, which incident group attributed to operational overload. In this noncredible video, it was alleged that USAID had provided millions of dollars to Hollywood Celebrities to fund their trips to Ukraine, where they allegedly met with president Vladimir Zelensky. This video went viral after it was reposted by several prominent accounts, including the owner of x, Elon Musk, as well as well as Donald Trump junior and the former attorney of president Trump, Sidney Powell. As of this writing, copies of this video have almost certainly, views of this, video have almost certainly passed, millions of engagements, including likes, reposts, and general comments in addition to views. So this was a a breakout influence event, almost certainly the most successful instance of this operation to date, and certainly a big feather in the cap of the individuals behind operation overload. Awesome. Thank you, Matt. Really, really poignant example there, so thanks for for running through that. Shifting gears a bit. So, so far we've centered the conversation and what we observed throughout 2024 into the beginning of 2025. I now wanna pose a question to all of you about what we expect to occur throughout the remainder of 2025. So what is an emerging threat, maybe one that we we started we saw, you know, begin to occur throughout 2024 into the beginning of the year that you think that we'll be talking about as as the year goes on. So, Derek, I'd I'd love to start with you. What's an emerging threat that you think that we'll be discussing when we, you know, run the same webinar, for for next year? Yeah. Thanks for the question, Maggie. So from my perspective and this kind of was set up a little bit, in my first answer in terms of what we saw in 2024, the greater, prominence of AI in threat actor tooling and and workflows. My answer here is more of the same, but on a much more sophisticated level. Right? And and what I mean by that is in in 2025, I would anticipate at some point seeing some sort of decentralized, semi autonomous, AI driven kill chain. And what I mean by that is not some sort of super robot that's going around, hacking everything in sight. But, what you can do is, you can have sort of cooperative agents sort of behind a what I would call a, like, a decentralized botnet or multiple decentralized botnets that are all responsible for an individual portion of the kill chain. Right? Or whatever framework you're using that can sequentially accomplish adversarial objectives. Right? We've seen individual components of it throughout 2024 and even in 2023. Right? We've seen, proofs of concept come out from universities where you have an AI agent that is able with I think it was, like, 87%, success rate able to, given a vulnerability in a piece of software, develop an exploit for that vulnerability. You have AI that are able to, you could have one set up using, Nmap or similar software to start doing reconnaissance and network, targeting, etcetera. And as each AI finds something within the target profile that it's been trained to find, it then chains itself to the next AI and says, okay. Here's your valid target. Move on with this. Then it says, okay. Cool. We've validated and, you know, accomplished our next part of the objective. Let's move it on to this next piece. Let me move it on to this next piece. Right? And in that, having those sort of different pieces all chained together, I think that's gonna be our next, evolution of the use of Gen AI in threat after, adversarial activity. Amazing. Thanks for for walking that through. So it sounds like, Derek, AI, very, very top of mind for you both in in what we saw throughout 2024, but but also looking forward as well. Thank you for for that answer. Matt, I'd I'd love to kick it to you. What do you see it as an and as as an emerging threat that we'll be talking about throughout the the rest of the year? Yeah. I really think over the course of 2025, we're gonna see China take increasingly coercive and provocative actions against against Taiwan. Although we still assess that an invasion before 2027 is unlikely, But these actions will likely include increasingly complex and large scale military exercises surrounding Taiwan, more aggressive and large scale, interdictions of Chinese of Taiwanese vessels and outlying waters, repeated severings of submarine cables, connected to Taiwan, as well as cyber incursions and attacks on Taiwan's energy, transportation, or financial sectors. Though Beijing's perceptions of how far it compressed Taiwan will very likely be colored by chaotic and at times competing policy making in Washington, DC, where you see a divide between China hawks and trade and and business focused elements. Beijing is also undoubtedly watching closely the unfolding negotiations to end Russia's war against Ukraine, most notably president Trump's prioritization of accruing financial benefit for The US versus punishing Russia's destabilizing actions on Russia on Europe's eastern border. Amazing. Thank you, Matt. And and to the questions coming in, it looks kinda based on that. So excited to to jump into those when we move to to the q and a portion. And then, Alex, what about you? What what's an emerging threat that that will be top of mind throughout the rest of 2025? Yeah. So for me and for for Insta Group in general, something that we're really interested in as we move into 2025 is the cybercriminal supply chain or the periphery surrounding cybercrime. So this goes along with the theme of decentralization earlier and what I said, in reference to the decline of as a service business models. Everything in cybercrime is interconnected, especially if we take ransomware as a case study. Ransomware groups, as elucidated by the concierge Blackbosta chat leaks, are collaborating openly with info stealer developers, bulletproof hosting providers, trafficker teams, scam groups, and so forth, so on into infinity, even some groups collaborating with, pro Russian hacktivists, and and and things like that. So as we move into 2025, as this becomes more fragmented, more splintered, and we have a a harder idea of understanding where the center of gravity is in cybercrime, we're going to have to transition to targeting the enablers of cybercrime. So looking at the infostealer operators, in particular, the infostealer developers and the trapper teams that are responsible for the global proliferation of malware, looking at those teams. And and we've done a ton of reports on groups like Marco Polo and Crazy Evil over the last six months trying to dig into this issue. Let's look at bulletproof hosting providers. Z servers was just taken down, obviously, last week and sanctioned, the enablers of infrastructure. Right? DDoS as a service providers that engage in double and triple extortion, the initial access brokers, the data leak brokers, even the forums themselves. We saw cracked and nulled went down earlier this month. I think more forums are going to be coming in 2025, but where does everyone go once those forums go down? Focusing on the enablers, the infrastructure, and the entire supply chain instead of just focusing on the end goal, the end objective, which is ransomware, that will able that that's gonna be better equipping us in order to fight the ransomware threats or any other cybercriminal threat as we move later into this year. Amazing. And thank you so much, Alex. I I did wanna ask them a follow-up question, one that we saw come in through the q and a. Obviously, a lot of shift in in where threat actors are are conducting conversations, where they're posting victims, and where are the conversations happening now after some of the takedown activity that we've observed throughout the year and and kind of the shifting generally in in the use of those forums? Yes. So we see, so so we break this down really into two different, centers of gravity, the English speaking and the Russian speaking cybercriminal underground. In the English speaking cybercriminal underground, especially with the takedowns of Cracked, Nulled, and the arrest of Pavel Dierov, the the CEO of Telegram, we are seeing a pivot towards gaming platforms, particularly Discord. We are also seeing a pivot towards social media. We are then on the Russian side beginning to see a pivot away from Telegram, which was a a a dominant, method of communication for the Russian speaking and and threat actors throughout the Commonwealth of Independent States. We're seeing a movement towards talks, key base, and other forms of end to end encrypted chat session, matrix, and so forth, movement away from public facing or semipublic facing communications into almost exclusively in group communications, where the only outreach those members of that in group have are to other service providers. So it we're we're seeing a transition away from openly announcing your activities on the forums, ransomware groups moving away from disclosing victims on a leak site. All of this has to do, again, with that theme of decentralization and that scattering. Threat actors don't trust the integrity and the privacy of these platforms any longer. They have to assume that most of these forums and special access sources are compromised in some capacity. Therefore, they're going to move, even more underground. Awesome. Thank you so much for for that in-depth description. Really appreciate it, Alex. So we've obviously, so far in the conversation, been talking about, you know, what what malicious actors are doing and how they're succeeding. I did wanna to end on a a kind of high note before we we move to q and a. Meaghan, from from your perspective, what do you see as a a positive trend, maybe a way in which we as defenders have succeeded over the past year? You know, what's a what's a highlight, that came about, in your research putting together this report? Yeah. Absolutely. Happy to to throw in a dose of optimism. I think that what we saw with the, TTP analysis with the shift to defense and discovery is a really good sign that defenders are doing something right. Because if you're a threat actor, you're not investing all of this time into defensive evasion. You're not looking at EDR evasion specific techniques. You're not trying to discover what's on the landscape if there was an easier way to do it. So I think that we're forcing them to adapt and to get around, some pretty baseline cyber defenses. And if more groups are doing that, that means that the low hanging fruit is actually getting a bit higher, which is good for the entire cyber cyber ecosystem. Right? Because we've seen a lot of defenses where a more poorly defended organization, which is connected to a highly defended organization, allows a chain of compromises that way. So I think that that is a good sign that, that the ecosystem cyber hygiene is improving. Awesome. Now back to the news. Well, hopefully not. I I did wanna pose, Megan, one of the questions that came in through the q and a to you. We had a question from Dawn about the efficacy of geofencing, and hard OBMFA. So I I wanted to pass that to you for for your thoughts. How how effective are some of those controls in reducing the exposure to credential theft? Yeah. Absolutely. I mean, those are both two very effective practices. If they would have been in place for the hundred and sixty five snowflake victims, for example, they would not be on that short list. So I think that what threat actors are doing is taking advantage of complexity in these systems that are made up with a lot of different as a service, cloud platforms and software that makes it difficult for defenders to kind of see the enterprise as a whole. So I think what that is a good argument for is a Swiss cheese version of defense where you add in your MFA, you add in your geofencing, you add in, network segmentation, all of which to try to make it increasingly difficult for threat actors to achieve their objectives without making it impossible for your users to get anything done. So, it's a constant evolution, I think, but those would two would be very effective. Awesome. Thank you, Megan. Derek, anything to add to to Megan's answer on that question about the efficacy of some of those those controls? Yeah. No. That's, that was a really good answer. I think, the only thing I would add here too is the pretty typical, threat intel analyst's answer of it depends. Right? And I'll be better and actually explain than usual. Right? So I think it depends on your configuration. It depends on your MFA. Right? And it depends on how strict your geofencing is. Not all MFA is created the same, and not all, sort of detection regimes are the same. So a lot of times what it's going to come down to are the robustness and thoroughness with which you apply your MFA across your authentication workflows. And it's also going to come down to signal convolution. Right? When you see your impossible travel or your impossible device shift, sometimes that can be explainable if it's one thing or one other thing. Right? But really what's critical when you're take when you're talking about account takeover and the use of valid accounts and and things like that is signal convolution. Is it multiple things changing rapidly over a short period of time? Are is that then coupled with behavioral shifts within the account itself? And so it's kind of you you kinda gotta go back to that defense in-depth type of a strategy where you're talking, hey. Yes. We can see these different, pieces, when you have your, you know, two FA, MFA, you know, defeat, but then also you need to kind of look at where that goes from there. And and, yeah, the MFA, like I said, kind of on a continuum. Right? You have your two FA, you have your MFA, you have your hard tokens, you have your Fido two. Right? And and depending on your network and and your, kind of, authentication workflow, really, the the thermos there makes the difference. Awesome. Thanks for for walking through that, Derek and Meaghan. Really appreciate the the perspective. Matt, I wanted to pivot back to you. We had a question come in as you were you were talking about the cable drag in Taiwan today. So I wanted to get your thoughts on on how, that maybe changes or or doesn't change into the assessments that you were speaking to. Well, I'd like to, make sure that to be clear that, I I wrote that assessment last week, so certainly before the event today. I'd like to get credit for that. I I think it really just, folds into what what we what I outlined in the assessment in terms of we we've seen submarine cable cuts, around Taiwan. It's more more frequently around Masu Islands, which are closer to Mainland China. But what we anticipate is there being more direct and aggressive actions around the Taiwan Central, so which isn't the case with this, this cable, severing recently. The the shipping question was, you know, import on Taiwan, and then the cable was severed, close to, Taiwan itself. So I think that's what you what we would anticipate seeing, increasingly, sort of, escalatory behavior on behalf of the Chinese. Awesome. Thanks, Matt. And, yeah, congrats. I guess, Alex, very quickly to to wrap us up before we close. We did have a question come in about some of the intel that was gleaned from the exposure of some of the black box to chat logs. So I wanted to toss that over to you before we wrap. Yeah. I mean, that is an entire webinar in and of itself. But a couple high level findings from that is, one, after three years, we finally have absolute confirmation that Blackbosta is a Conti Splinter group, based on the discussions within, not only the logs, but also the the general, gossiping that's taking place there. It's administration. It's tooling providers. We do we do obviously have confirmation there. Something really important is the highly specialized and highly organized nature of Blackbosta. We see individuals who speak English, who are specialized in social engineering. We see highly targeted operations, someone tasked with running through Zoom info just by themselves. That's their entire job. We have dedicated negotiators. We have dedicated infrastructure providers, points of contact and middlemen for Infostealer providers like Luma Steeler or a bunch of, tools and services like a a a Goblin Crypto, which has never been discovered before that has just been, kind of unearthed, unmasked from the back Blackbosta logs. So expect a lot more out of us, especially, customer facing in the coming days and weeks related to Blackbosta. But, it's going to take a long time for a lot of that information to come out. Last I checked, there was about 1,200,000 lines of chat logs in there. So so we're gonna have a a lot of of work ahead of us, but overall, legitimate and and very interesting, and and I can't wait to see, what people find in there. Amazing. Well, thank you so much, Alex, for that, and and excited for kinda more to come after some some more in-depth analysis there. I do wanna be cognizant of everyone's time and and note that we are at the top of the hour. Thank you so much for joining us for today's conversation about Recorded Futures 2024 annual threat landscape. I know that we've had lots of questions come in in the chat. To the extent that you have any further questions, please feel free to to reach out to the folks that you see here on the screen, or engage with Recorded Future in another way. We'd love to answer those questions, over the coming days and weeks. Thank you so much for your time, and, we'll hopefully chat soon. Bye, everybody. Have a great day.