DarkWeb Politics. Cybercrime & Meddling Threat Landscape
Updated: Oct 8
With the upcoming 2020 elections and increasingly heated political discussion, the threat of social media manipulation and election meddling employed by external actors is fast becoming a reality. more and more tangible. Indeed, social media platforms can be abused by applying automation, impersonation, and deliberate false information dissemination in order to alter organic socio-political discussion processes.
At the same time, the culturally contradictory and crucially sensitive nature of these malicious activities often results in a shift from the threat discussion per se towards political statements and social disputes. These disputes focus on nation-state actors and government-lead threat actor teams such as the Internet Research Agency, while the massive and extremely developed infrastructure of cybercrime and for-profit private threat actor groups often remains out of the discussion.
In this joint research by AdvIntel with the NatSec commentary by Nine Mile Security Group (NMSG), we focus on the non-state, “shadow” side of political meddling. We investigate the Tactics, Techniques, and Procedures (TTPs), infrastructure, and methods used by foreign, non-politically motivated threat actors who are willing to participate in illicit activities aimed to impact the US political environment.
Policy, Privacy, and Social Media (by Nine Mile Security Group)
The 2008 Presidential Election unlike any other before it contributed to the meteoric rise of social media being used for targeting, advertising and in some cases misrepresentation of information to United States citizens. Set against a backdrop of noticeable social and cultural changes that have either widened or highlighted long-standing divides, hackers, cyber-criminals and everyone else in between have raced to occupy the void with agendas of their own.
Longstanding and relatively safe topics like Education and to some extent Immigration have become both wedge and legitimate security issues. Whether the debate centers on assistance to low-income, inner-city-residing minorities, public loan forgiveness to borrowers often part of the same community, or pathways to citizenship to those seeking American opportunity, what has been clear is that the dissatisfaction of voters across the board for whom these issues are critically important has been weaponized and used to manufacture and often times provoke discord by bad actors.
The massive collection efforts by government and big business alike of innumerable data points have as of late appeared in starker relief to the general public. And now, data, the most valuable commodity on earth has also proven to be the least understood. According to a survey conducted by research firm IPSOS, sixty-six percent of people knew very little about how much data companies (and by extension hackers, cyber-criminals who later pilfer said data) held about them, or what was done with it. Additionally, only about thirty-three percent had a fair amount of trust in companies and the government using it in the right way. These statistics are part and parcel of what provides fertile ground for the social media-based disinformation campaigns cultivated and levied by bad actors for nefarious purposes.
For instance, the ‘social media trap’ set by foreign agencies has been met with widespread success in recent years. It essential is a mechanism by which people get trapped based on some action taken; likes and comments on their post by unknown people. If aimed at a military target, crucial information about logistics, operational details, training institutions, details of weapon ammunition, troop movements and other vital information is collected.
Politics, Business, and National Security
Data can be harvested from a cornucopia of sources for both political and business use. Some of the most well-known being from, web-browsers, ride-sharing platforms, peer-to-peer funds transfer apps, or the dark web itself. With a dearth of federal regulation that uniformly and explicitly addresses how political campaigns or businesses writ large ought to collect, use, and share data, there are little to no safeguards that effectively stymie and/or prevent the targeting and subsequent exploitation of the public for ill-suited purposes by bad actors.
Consider for a moment that during the 2016 Presidential Election (and for 2020) employees from top social media companies were embedded in both Republican and Democratic campaigns as “digital strategists”. While this in and of itself is one hundred percent part of our democratic process, it also to some extent widens one particular divide; specific tools used to create highly targeted ads sent via social media to an ideal voter or bloc of voters.
Sound familiar? Well, that is because while for completely opposite purposes, bad actors employ the very same tactics with the intent to further fracture an already polarized public.
Cyber-threats, nefarious activity, and attacks are top of mind within the business community. In a report where over one thousand businesses from Latin America, the Caribbean, Europe, United States, Canada, Asia Pacific, and Africa were surveyed on the security climate as part of a 2019 “Global Cyber Risk Perception” survey, these were some key takeaways:
Seventy-nine percent view cyber-attacks as a top-five priority
Only 17% of executive leaders and board members spent more than a few days focusing on cyber risk
Top two attack zones: Cloud Computing and IoT Devices
So, what does this all mean for the future of our collective security? Well, where you stand largely determines where you sit, but these few things are for sure:
The sheer financial cost of cybercrime is already in the billions of dollars – and rises year over year.
We need to understand who our adversaries are, where they’re located, and what their capabilities, plans, and intentions are.
The dividing lines are at best fluid, public debate is fever-pitched, and threat actors are fast gaining ground.
Time is of the essence; we truly need all hands on deck.
Research Methodology: Ethics & Safety First
We believe that it is essential to disclose the key points of the research and the methodology used in this investigation due to the extreme sensitivity of the discussed issue.
In order to build a comprehensive understanding of the threat landscape, Advintel relied on the adversary perspective, meaning that we investigated tactics shared by the actors themselves. We initiated threat actor engagements with over 40 foreign cybercriminals and professional meddlers who emphasized their willingness to play a part in a political manipulation activity. At the core of this group were high-profile threat actors previously tracked by our Subject Matter Experts (SME) team who had a proven record of active participation in elite cybercrime communities.
We declined all offers from threat actors to conduct proof-of-concept/proof-of-value (POC/POV) operations. These operations were offered to be conducted in a non-US social environment; however, AdvIntel believes that disinformation activities even conducted on a minor scale for research purposes are deeply unethical and immoral. Additionally, we declined any offers to create samples POC/POV materials such as political content writing, video creation or any other artifacts which could have been used in an actual manipulation campaign, after our investigation was finished.
Even though the strategies shared by the threat actors emphasized teamwork and tight cooperation between different groups of cybercriminals involved in a campaign, we have not introduced the actors to each other or created any joint chats or spaces in which the investigated actors could have communicated. This was done to avoid creating social capital for threat actors which could be used in future malicious activities.
As our main focus was on an adversary perspective, we relied on actors sharing strategies themselves. As a result, the majority of actors asked for specific details about the political environment in which they would need to operate to develop the most particular and efficient ways of manipulating this environment. We were able to convince the actors to model their approaches based on the most abstract case possible - a non-existent US state with a population of 10-15 million people going through Congressional elections. We have deliberately avoided any modeling based on real US communities, jurisdictions, or locations.
Finally, while designing the investigation we have consulted with security researchers, and academic experts to ensure that even indirect facilitation of any damage to the US social, cultural or political fabric.
Step 1. DarkWeb Politics
To assess the threat landscape we asked a simple question - what will it take for a foreign malicious group to impact the US political environment through illegal means?
To answer this question we went to space which we, as security researchers, knew the best - the DarkWeb and criminal underground. Later in our investigation, we discovered that threat actor participating in meddling would have done the same. In an interview with a source, based in Eastern Europe who claimed to actively participate in defamation and disinformation campaigns during local regional elections, the source argued that approaching the DarkWeb is always their first step. DarkWeb provides an already established infrastructure of established connections between people, an integrated network of specialists who value their reputation. This enables anonymously to hire experts with minimizing the risks of been scammed or detected. Moreover, the networks of escrows and moderators which monitor the deals within the community allow ensuring that the hired actors will not disappear with the money.
To fully utilize the benefits of the DarkWeb trust sharing networks we selected several top-tier cybercrime forums and job sites. These forums were entirely devoted to hacking, carding, financial fraud and other types of cybercrime. Political discussions, hacktivism, and other actions were prohibited by the administrators and moderators, which ideally fit our research goals to test how active and efficient cybercrime syndicates can be when it comes to operations of political nature.
As a result, in less than two weeks and spending $0 we were able to assemble a pool of 40 experienced cyber criminals who had a coherent strategy and were eager to start the operation “the next day.” Three findings were particularly important about this group.
First, many actors were of an international background. Our engagements were done through Russian-speaking forums, which was defined exclusively by two reasons - this was the environment in which we could have the most accurate threat assessment and the most developed analysis, secondly, these communities are home to the most sophisticated and experienced crime groups and actors.
Therefore, naturally, a large number of actors interviewed were Russian-speakers (which does not suggest that these actors were ethnically Russian or Russian-based). However, we have observed a solid number of non-Russian participants. The actors engaged claimed to be based in China, the MENA region, the EU, South Asia, and Oceania. We assess with a high level of confidence that the international actors are attracted to the infrastructural and network capabilities of Russian-speaking top-tier forums.
Secondly, criminals who demonstrated an interest in meddling activities came from all possible domains of cybercrime. This means that when it comes to cyber-political manipulation the threat environment is extremely malleable - methods used by ransomware groups, carders, and traffic hosters can be creatively weaponized for the sake of intrusive social actions into the US environments.
Third, despite the for-profit nature of forums, some of the actors actually did have political experience. This group primarily focused on blackhat search engine optimization (SEO) and traffic herding to boost illegal content. These actors were arguing that their experience tested within foreign environments can be successfully transferred into a US setting.
Step 2. The Meddling Instruments
As the main investigative strategy was the discovery of the adversarial perspective, the actors themselves were convinced to actively share strategies and visions for the targeted goal. All of these suggestions can be categorized as an adaptation of cybercrime to political needs, specifically, carding and financial fraud, ransomware and network intrusions, and web fraud and traffic manipulations.
Carding, also known as card fraud is one of the major components of the modern cybercrime ecosystem. The carder skills are seemingly distant from politics; however, they can be perfectly adapted for the purposes of political meddling operations.
According to the threat actors, the first thing required to successfully spread disinformation over social media is a system of impersonated accounts that seem authentic while bypassing social media platform algorithms. This process is almost identical to online credit card fraud. When carders make a purchase using a stolen card they should bypass anti-fraud by acting exactly as the cardholder. This sophisticated process requires obfuscation of real digital fingerprints, the use of anti-detect frameworks, and complex social engineering. An experienced carder can operate a stolen US bank account, mimicking its real owner and remain undetected. The same skills can be used to steal an account or create a sockpuppet on social media and successfully operate it to disseminate political content. In addition to this, carders offer to use their existing network of impersonated financial accounts to manage all the payments needed for the meddling operation.
Moreover, just as political campaigns sometimes require physical presence, carding operations have a physical dimension - so-called drop services. Each developed carding enterprise has a team of people who will intercept stolen packages, verify purchases and handling the situation when a physical presence of the impersonated account owner is needed. Drop services already have teams on the ground with an established infrastructure of fake identities, IDs, warehouses, and vehicles. These teams can be used for the physical implementation of disinformation - placement of banners, flyers, cold-calling, or even provocations.
According to the actors, malware and ransomware can be especially useful when it comes to sabotaging activities. Well-organized malware syndicates make a living by breaching into secure networks, including government, defense contractors, financial institutions, and healthcare. Environment belonging to political organizations are not always as protected as some of these groups’ top targets. Syndicates often rely on the best credential stealer crimeware tool which enables them to receive login credentials and access correspondence and financial records.
Targeted breaches into the network may result in correspondence exposure and other crucial reputational damage or fatal business interruption caused by ransomware. Even if the breach team is unable to access the network or will be unable to steal information, they may report a fake breach through top-tier underground forums, while an affiliated social media team will spread the misinformation about the cyber attack.
Finally, web traffic manipulation can easily be used for propaganda purposes. Different methods of search engine manipulation and traffic redirect allows altering the search result order for a targeted request, source, or website. Redirected traffic from obscene sources such as pornography websites can be used to compromise the legitimate web source rate standings, while redirect pages can be used to increase the visibility of provocative content.
Step 3. From Stealing Money to Weaponizing Narratives
To understand the application of cybercrime to propaganda in a more exact way, we have accumulated different strategical designs offered by threat actors. The most efficient solutions emphasized the creation of a work process that will organically embed all different methods in a single campaign, with each methodology amplifying the other. As a result, the final strategic view presumed the organization of the forty engaged actors into four Groups.
Group 1. Political Content Writers. This team would include experienced content writers and social media promotion experts who had previous experience in political meddling. In our pool of actors, we have identified six people from the EU and Eastern Europe who were willing to take these responsibilities and design a political engagement strategy. The most developed versions focused on the following methods centered around active tribalization and alienation of the political process:
Step 1. Identification of dividing content to amplify the feeling of political apathy. The types of content included issues of racial divides, environmental issues, ethnic and confessional divides, immigration, crime rates, police brutality, unemployment, urban-/rural divide, utility pricing, and housing crisis. It is noteworthy that even though the actor did mention racial and religious divides, their main emphasis was made on the socio-economic rather than identity hotspots.
Step 2. Group 1 will identify the already-existing political content produced by radical and extremist proponents of ideologies related to dividing issues. This content will become the foundation of the media campaign, being actively disseminated to disproportionately invade the public discourse with radical ideas, and, this way, delegitimize it.
Step 3. Simultaneously new content is being produced (mostly via basic digital means) which will entirely focus on out-of-context controversial statements made by legitimate politicians, portraying them as a proponent of a certain extremist ideology and exclusive advocates of only a selected voter group. This way the complex and multidimensional political agenda of legitimate candidates will be reduced, exacerbating the social divide, while alienation between voter groups will increase. Special negative targeting will be aimed at moderate, reconciliatory, and universal-values-based political narratives that are designed to address the needs of all voters within the state. The threat actors emphasize that after a certain point, each message will start spreading on its own creating a chain reaction due to an already-high alienation and clusterization of the US society.
Step 4. In an atmosphere of mistrust and apathy, one of the political groups convenient for the meddling organizers is being selected and mobilized. Mobilization efforts are also conducted through social media pages.
These measures are aimed to decrease the voting rate, by spreading perception of the candidates themselves as manipulators serving in the interest of selected social groups and pursue narrow political agendas. With the lower voting rate the chances of the meddling increase due to the lowering of amortization. If the meddlers are interested in a radical agenda to prevail, a small radical but very vocal and mobilized political group that would otherwise receive marginal results (in absolute numbers) may now receive high results in relative percentage numbers, as the overall number of voters is low. In case, the meddlers are on the opposite are interested in a moderate candidate, the erosion of political discourse can compel voters to align with the most rational candidate during the overall panic.
As we can see this approach bears similarities to the one attributed to the Internet Research Agency, however, it has a more coherent and logical inner structure. Whereas the evidence available regarding the Agency’s operations suggest that the troll farm was massively and relatively randomly spreading provocative messaging to different segments of the US society, the paradigm taken by the for-profit threat actors bears a more pragmatic consistent approach. In their strategy, all the steps are tightly connected one with another and all serve to achieve a clear final goal - the benefits of a selected political agenda. This is most likely related to the fact that the main experience of these actors was promotion and advertisement (mostly of illicit content and commodities) and they approach political meddling as a marketing campaign.
It is noteworthy that fake news per se are not widely mentioned by the actors, as direct disinformation is found to be inefficient since it triggers account bans. Instead, they prefer to focus on the content which does not directly misinform the audience, by rather shift the emphasis in a disproportional way, facilitating the erosion of political culture and minimizing the opportunity for a discussion or a search for common ground emerging between opposite political views.