Category Archives: medium

Securelist: Remotely controlled EV home chargers – the threats and vulnerabilities

We are now seeing signs of a possible shift in the field of personal transport. Recent events such as the ‘dieselgate’ scandal undermine customer and government confidence in combustion engines and their environmental safety. At the same time there has been a big step forward in the development of electric vehicles. In addition to favorable media coverage, modern EVs have evolved a lot in terms of battery endurance, driving speeds and interior and exterior design.

To stimulate growth in the personal EV segment some countries even have special tax relief programs for EV owners. But there is still a major problem – the lack of charging infrastructure. This may not be as relevant in big cities, but in other places car owners mostly rely on their own home EV chargers, a relatively new class of device that has attracted our attention.

There are lots of home charger vendors. Some of them, such as ABB or GE, are well-known brands, but some smaller companies have to add ‘bells and whistles’ to their products to attract customers. One of the most obvious and popular options in this respect is remote control of the charging process. But from our point of view this sort of improvement can make chargers an easy target for a variety of attacks. To prove it we decided to take one of them, ChargePoint Home made by ChargePoint, Inc., and conduct some in-depth security research.

ChargePoint Home supports both Wi-Fi and Bluetooth wireless technologies. The end user can remotely control the charging process with a mobile application available for both iOS and Android platforms. All that’s needed is to register a new account in the application, connect a smartphone to the device via Bluetooth, set the parameters of a Wi-Fi network for an internet connection, and finish the registration process by sending the created user ID and the smartphone’s GPS coordinates to the backend from the device.

In a registered state, the device establishes a connection to the remote backend server, which is used to transfer the user’s commands from the application. The application thereby makes it possible to remotely change the maximum consumable current and to start and stop the charging process.

To explore the registration data flows in more detail, we used a rooted smartphone with the hcidump application installed. With this application, we were able to make a dump of the whole registration process, which can later be viewed in Wireshark.

The Bluetooth interface is only used during the registration phase and disabled afterwards. But we found another, rather unusual wireless communication channel that is implemented by means of photodiode on the device side and photoflash on the smartphone side. It seems to have just one purpose: by playing a special blinking pattern on the flash, the application can trigger the factory reset process after the device’s next reboot. During the reboot, Wi-Fi settings and registered user information will be wiped.

In addition, we found a web server with enabled CGI on the device. All web server communications are protected by the SSL protocol with the same scheme as the control server, so the web server inherits the described certificate security issue. We discovered a series of vulnerabilities in CGI binaries that can be used by an intruder to gain control of the device. Two of them were found in the binary used to upload files in different folders to the device depending on the query string parameters. Other vulnerabilities (stack buffer overflow) were found in the binary used to send different commands to the charger in the vendor-specific format (included in a POST message body). We also found the same stack buffer overflow vulnerabilities in the other binary used for downloading different system logs from the device. All this presents attackers with an opportunity to control the charging process by connecting to the target’s Wi-Fi network.

Vulnerabilities in the Bluetooth stack were also found, but they are all minor due to the limited use of Bluetooth during regular device operation.

We can see two major capabilities an intruder can gain from a successful attack. They will be able to:

  • Adjust the maximum current that can be consumed during charging. As a result, an attacker can temporarily disable parts of the user’s home electrical system or even cause physical damage – for example, if the device is not connected properly, a fire could start due to wires overheating.
  • Stop a car’s charging process at any time, for example, restricting an EV owner’s ability to drive where they need to, and even cause financial losses.

We sent all our findings to ChargePoint, Inc. The vulnerabilities we discovered have already been patched, but the question remains as to whether there is any reason to implement wireless interfaces when there is no real need for them. The benefits they bring are often outweighed by the security risks they add.

Download “ChargePoint Home security research” (English, PDF)



Securelist

Securelist: Zero-day in Windows Kernel Transaction Manager (CVE-2018-8611)

Executive summary

In October 2018, our AEP (Automatic Exploit Prevention) systems detected an attempt to exploit a vulnerability in the Microsoft Windows operating system. Further analysis led us to uncover a zero-day vulnerability in ntoskrnl.exe. We reported it to Microsoft on October 29, 2018. The company confirmed the vulnerability and assigned it CVE-2018-8611. Microsoft just released a patch, part of its December update, crediting Kaspersky Lab researchers Boris Larin (Oct0xor) and Igor Soumenkov (2igosha) with the discovery.

This is the third consecutive exploited Local Privilege Escalation vulnerability in Windows we discovered this autumn using our technologies. Unlike the previously reported vulnerabilities in win32k.sys (CVE-2018-8589 and CVE-2018-8453), CVE-2018-8611 is an especially dangerous threat – a vulnerability in the Kernel Transaction Manager driver. It can also be used to escape the sandbox in modern web browsers, including Chrome and Edge, since syscall filtering mitigations do not apply to ntoskrnl.exe system calls.

Just like with CVE-2018-8589, we believe this exploit is used by several threat actors including, but possibly not limited to, FruityArmor and SandCat. While FruityArmor is known to have used zero-days before, SandCat is a new APT we discovered only recently. In addition to this zero-day and CHAINSHOT, SandCat also uses the FinFisher / FinSpy framework.

Kaspersky Lab products detected this exploit proactively through the following technologies:

  1. Behavioral detection engine and Automatic Exploit Prevention for endpoint products
  2. Advanced Sandboxing and Anti Malware engine for Kaspersky Anti Targeted Attack Platform (KATA)

Kaspersky Lab verdicts for the artifacts used in this and related attacks are:

  • HEUR:Exploit.Win32.Generic
  • HEUR:Trojan.Win32.Generic
  • PDM:Exploit.Win32.Generic

Brief details – CVE-2018-8611 vulnerability

CVE-2018-8611 is a race condition that is present in the Kernel Transaction Manager due to improper processing of transacted file operations in kernel mode.

This vulnerability successfully bypasses modern process mitigation policies, such as Win32k System call Filtering that is used, among others, in the Microsoft Edge Sandbox and the Win32k Lockdown Policy employed in the Google Chrome Sandbox. Combined with a compromised renderer process, for example, this vulnerability can lead to a full Remote Command Execution exploit chain in the latest state-of-the-art web-browsers.

We have found multiple builds of exploit for this vulnerability. The latest build includes changes to reflect the latest versions of the Windows OS.

Check for the newest at the moment Windows 10 Redstone 4 Build 17133

A check for the latest build at the time of discovery: Windows 10 Redstone 4 Build 17133

Similarly to CHAINSHOT, this exploit heavily relies on the use of C++ exception handling mechanisms with custom error codes.

To abuse this vulnerability exploit first creates a named pipe and opens it for read and write. Then it creates a pair of new transaction manager objects, resource manager objects, transaction objects and creates a big number of enlistment objects for what we will call “Transaction #2”. Enlistment is a special object that is used for association between a transaction and a resource manager. When the transaction state changes associated resource manager is notified by the KTM. After that it creates one more enlistment object only now it does so for “Transaction #1” and commits all the changes made during this transaction.
After all the initial preparations have been made exploit proceeds to the second part of vulnerability trigger. It creates multiple threads and binds them to a single CPU core. One of created threads calls NtQueryInformationResourceManager in a loop, while second thread tries to execute NtRecoverResourceManager once. But the vulnerability itself is triggered in the third thread. This thread uses a trick of execution NtQueryInformationThread to obtain information on the latest executed syscall for the second thread. Successful execution of NtRecoverResourceManager will mean that race condition has occurred and further execution of WriteFile on previously created named pipe will lead to memory corruption.


Proof of concept: execution of WriteFile with buffer set to 0x41

As always, we provided Microsoft with a proof of concept for this vulnerability, along with source code. And it was later shared through Microsoft Active Protections Program (MAPP).

More information about SandCat, FruityArmor and CVE-2018-8611 is available to customers of Kaspersky Intelligence Reports. Contact: intelreports@kaspersky.com



Securelist

Kaspersky Lab official blog: Online fraud: 5 most common spammer tricks

Spam and phishing often go hand in hand: Fraudsters send mass mailings in an attempt to phish information from recipients. For them, users’ personal data remains a highly prized and desired asset, as evidenced by both the constant high-profile media stories and our own spam flow analysis. A common aim of spam is to gain access to your accounts or bank card numbers through e-mail phishing and social engineering techniques.

Online fraud: 5 most common spammer tricks

In this post, we look at the five tricks most commonly employed by spammers.

1. Fake notifications from social networks

Spammers actively send fake notifications that seem to come from popular social networks and are about new friends, their activities, comments, likes, and so forth. Such messages are often indistinguishable from the real thing, the only difference being that they contain a phishing link, which is not always easy to spot. On following the link, users are prompted to enter their username and password on a fake login page.

Another common variant is messages supposedly from social networks, but this time with threats alleging, for example, that suspicious activity has been detected on your account, or that a new feature has been introduced and users who don’t give their consent will be blocked. Whatever the case, the message will contain a button with a link to a phishing login page.

Phishers most popular tricks: Fake notifications from social networks

Phishers most popular tricks: Fake notifications from social networks

2. Banking phishing

Phishing aimed at obtaining users’ bank card details is still the most popular kind of fraud. Fake messages may be sent in the name of banks or payment systems. The most common message subjects are related to account blocking or “suspicious activity” on the client’s personal account.

Under the pretext of restoring access, confirming identity, or canceling a transaction, the user is asked to enter bank card details (often including the CVV/CVC code) on a fake bank website. On receiving the data, the criminals immediately withdraw money from the victim’s account. It’s the same story with payment systems, but in those cases, victims are asked only to log in to their accounts.

Phishers most popular tricks: Fake notifications from banks and payment systems

Phishers most popular tricks: Fake notifications from banks and payment systems

3. Fake notifications from popular services and sellers

Likewise, fake notifications are created using the brand names of popular online stores, delivery services, booking sites, multimedia platforms, job search websites, and other popular online services. Cybercriminals rely on the odds their spam messages will reach at least some genuine users of such services, who are likely to panic and click or tap whatever they see.

Phishers most popular tricks: Fake notifications from various services and shops

Phishers most popular tricks: Fake notifications from various services and shops

4. Fake notifications from e-mail services

Scammers use this kind of spam to harvest usernames and passwords for e-mail services. One of two common pretexts is typically deployed: Users are prompted either to restore their password or to increase the available space in their mailbox, which is supposedly full. In the latter case, the phishing link promises a manifold increase in storage capacity, which in the era of cloud computing and the ever-growing need for storing large amounts of data does not seem all that suspicious.

Phishers most popular tricks: Fake notifications from e-mail services

Phishers most popular tricks: Fake notifications from e-mail services

5. “Nigerian prince” fraud

Lastly, one of the oldest types of spam — the promise of fortune from a relative or a lawyer of a dead millionaire in exchange for an up-front payment — is still making the rounds. A variation on the theme involves the scammer posing as a celebrity in a difficult situation. Victims are promised an impressive reward if they agree to help the unfortunate millionaire withdraw funds trapped in various bank accounts. To do so, they must, of course, first send detailed information about themselves (passport details, account data, etc.) and a modest amount of money for paperwork.

Phishers most popular tricks: Nigerian prince fraud

Phishers most popular tricks: Nigerian prince fraud

The list of spammers’ favorite topics and techniques does not end there, but the five methods described above are the most effective and thus the most common.

Don’t be a victim

The best advice is to be careful. But that’s a bit vague, so here’s some nitty-gritty:

  • When you receive a message with a notification from a company or service, check that it was sent from a bona fide address. Using Google as an example, the message should come from no-reply@accounts.google.com, and not no-reply@accounts.google.scroogle.com or something like that.
  • If you do follow a link in such a message, again, make sure that you are taken to the real website, not a fake.
  • Use a reliable security solution with antispam and antiphishing protection — it will detect fraudulent e-mails and warn you clearly.


Kaspersky Lab official blog

Securelist: DarkVishnya: Banks attacked through direct connection to local network

While novice attackers, imitating the protagonists of the U.S. drama Mr. Robot, leave USB flash drives lying around parking lots in the hope that an employee from the target company picks one up and plugs it in at the workplace, more experienced cybercriminals prefer not to rely on chance. In 2017-2018, Kaspersky Lab specialists were invited to research a series of cybertheft incidents. Each attack had a common springboard: an unknown device directly connected to the company’s local network. In some cases, it was the central office, in others a regional office, sometimes located in another country. At least eight banks in Eastern Europe were the targets of the attacks (collectively nicknamed DarkVishnya), which caused damage estimated in the tens of millions of dollars.

Each attack can be divided into several identical stages. At the first stage, a cybercriminal entered the organization’s building under the guise of a courier, job seeker, etc., and connected a device to the local network, for example, in one of the meeting rooms. Where possible, the device was hidden or blended into the surroundings, so as not to arouse suspicion.

High-tech tables with sockets are great for planting hidden devices

High-tech tables with sockets are great for planting hidden devices

The devices used in the DarkVishnya attacks varied in accordance with the cybercriminals’ abilities and personal preferences. In the cases we researched, it was one of three tools:

  • netbook or inexpensive laptop
  • Raspberry Pi computer
  • Bash Bunny, a special tool for carrying out USB attacks

Inside the local network, the device appeared as an unknown computer, an external flash drive, or even a keyboard. Combined with the fact that Bash Bunny is comparable in size to a USB flash drive, this seriously complicated the search for the entry point. Remote access to the planted device was via a built-in or USB-connected GPRS/3G/LTE modem.

At the second stage, the attackers remotely connected to the device and scanned the local network seeking to gain access to public shared folders, web servers, and any other open resources. The aim was to harvest information about the network, above all, servers and workstations used for making payments. At the same time, the attackers tried to brute-force or sniff login data for such machines. To overcome the firewall restrictions, they planted shellcodes with local TCP servers. If the firewall blocked access from one segment of the network to another, but allowed a reverse connection, the attackers used a different payload to build tunnels.

Having succeeded, the cybercriminals proceeded to stage three. Here they logged into the target system and used remote access software to retain access. Next, malicious services created using msfvenom were started on the compromised computer. Because the hackers used fileless attacks and PowerShell, they were able to avoid whitelisting technologies and domain policies. If they encountered a whitelisting that could not be bypassed, or PowerShell was blocked on the target computer, the cybercriminals used impacket, and winexesvc.exe or psexec.exe to run executable files remotely.

Verdicts

not-a-virus.RemoteAdmin.Win32.DameWare
MEM:Trojan.Win32.Cometer
MEM:Trojan.Win32.Metasploit
Trojan.Multi.GenAutorunReg
HEUR:Trojan.Multi.Powecod
HEUR:Trojan.Win32.Betabanker.gen
not-a-virus:RemoteAdmin.Win64.WinExe
Trojan.Win32.Powershell
PDM:Trojan.Win32.CmdServ
Trojan.Win32.Agent.smbe
HEUR:Trojan.Multi.Powesta.b
HEUR:Trojan.Multi.Runner.j
not-a-virus.RemoteAdmin.Win32.PsExec

Shellcode listeners

tcp://0.0.0.0:5190
tcp://0.0.0.0:7900

Shellcode connects

tcp://10.**.*.***:4444
tcp://10.**.*.**:4445
tcp://10.**.*.**:31337

Shellcode pipes

\\.\xport
\\.\s-pipe



Securelist

Securelist: APT review of the year

What were the most interesting developments in terms of APT activity throughout the year and what can we learn from them?

Not an easy question to answer; everybody has partial visibility and it’s never possible to really understand the motivations of some attacks or the developments behind them. Still, with the benefit of hindsight, let’s try to approach the problem from different angles to get a better understanding of what went on.

On big actors

There are a few ‘traditional’ actors that are very well known to the security community and that everybody has been tracking for the last few years. It has been business as usual for these actors in 2018 or, if anything, perhaps slightly quieter than usual.

In reality, it is the doctrines and modi operandi of these groups that determine how they react in the event of their operations becoming public knowledge. Some actors will simply abort their campaign and go into clean-up mode, while others carry on as normal. In order to do so, it is common for some of these actors to simultaneously work on several sets of activity. This allows them to compartmentalize operations, and if they are discovered, they simply improve their toolset to avoid detection next time.

We traditionally find many Russian-speaking actors in this second group, and we would like to highlight the 2018 activity of Sofacy, Turla and CozyBear.

Sofacy was probably the most active of the three. Throughout the year we detected it in various operations, updating their toolset and being blamed by authorities for several past operations. We have seen the actor deploying Gamefish and an updated version of its DealersChoice framework against embassies and EU agencies. One of the most high-profile incidents was abuse of Computrace LoJack by this actor in order to deploy its malware on victim machines, in what can be considered a UEFI-type rootkit.

Zebrocy is one of the tools traditionally used by this actor, but in reality the collection of cases where this tool was used can be considered a subset of activity in its own right. We saw different improvements for Zebrocy’s subset, including a new custom collector/downloader, new VBA implementing anti-sandboxing techniques and new .NET modules.

During the year we understood that Sofacy appears to be changing at a structural level and is possibly already being split into different subgroups. With the OlympicDestroyer analysis we learnt that this highly sophisticated false flag operation was somehow related to Sofacy. However, we later observed more activity by the OlympicDestroyer subset in Europe and Ukraine, and it was then that we decided to treat it as the entity we call Hades.

Of particular interest is how, after the publication of the GreyEnergy set of activity that is believed to be a continuation of BlackEnergy/Sandworm, we found additional overlaps between GreyEnergy and Zebrocy, including the use of the same infrastructure and the same 0-day for ICS.

All that seems to link this new Hades actor with the Zebrocy subset of activity, traditionally attributed to Sofacy, as well as part of the BlackEnergy/GreyEnergy/Sandworm cluster.

Regarding Turla, we didn’t spot any big structural changes like those described above, though we did see this actor using some interesting implants such as LightNeuron (targeting Exchange servers as described in our previous APT summary for Q2), as well as a new backdoor that, according to ESET, infected Germany’s Federal Foreign Office in 2017, as well as other entities in the European Union.

We discovered this actor using a new variant of its Carbon malware in its traditional activity of targeting embassies and foreign affairs institutions throughout the year. It also started using a new framework that we call Phoenix, as well as (unsurprisingly) transitioning to scripting and open source tools for its lateral movement stage.

Finally, some potential CozyDuke activity was detected during November 2018, apparently targeting diplomatic and governmental entities in Europe. The TTPs do not seem to be those that are usually attributed to this actor, which opened the door to speculation about this malware being used by a different group. The facts still seem to confirm that the malware used is attributable to CozyDuke. We are still investigating this new campaign by an actor that has been inactive for months.

It’s also worth mentioning Lazarus and BlueNoroff activity in 2018. We observed constant activity from this group targeting different regions including Turkey, other parts of Asia and Latin America, as well as various lines of business that provide it with financial gain, such as casinos, financial institutions and cryptocurrencies. In its more recent campaigns it has started deploying a new malware we call ThreatNeedle.

On false flags

It comes as no surprise to find false flags every now and again, sometimes implemented rather naively. But this year we witnessed what should be considered (so far) the mother of all false flags (more details can be found here). Other than the technical details themselves, what is also worth considering is the real purpose of this attack, and why these sophisticated false flags were planted in the malware.

The first obvious conclusion is that attackers now understand very well what techniques are used by the security industry to attribute attacks, so they have abused that knowledge to fool security researchers. Another consideration is that the main objective of an attack is not necessarily related to stealing information or disrupting operations – imitating an attacker might be more important.

This may actually be part of what some actors are doing at the moment. There are several groups that were apparently inactive for some time but now appear to be back. However, they are using different TTPs that are not necessarily better. As we shall see later, a couple of examples may be CozyDuke and APT10. As a purely speculative thought, it might be that their traditional toolset is now being used by different groups, maybe still related to the original operators. The purpose might be to make attribution more difficult in the future, or simply to distract from their real ongoing operations.

The whole OlympicDestroyer story eventually resulted in the discovery of a new subset of activity related to both Sofacy and BlackEnergy that we call Hades. We will see how these more sophisticated false flags evolve in the future and how they are used to pursue less explicit goals.

On the forgotten ones

Throughout the year we also saw how several old ‘friends’ re-emerged from hibernation with new sets of activity. Here we are talking about several well-known actors that for unknown reasons (a lack of visibility might be one of them) didn’t display much activity in recent times. However, it seems they are back. In some cases they appear in different weaker forms, perhaps with different operators, or just pretending not to be in shape while they run other parallel operations; in others cases they are back with their usual capabilities.

We can summarize all this by dividing it up into the regions that showed most activity during the year. First place went to South East Asia, followed by the Middle East.

For South East Asia we can point to groups such as Kimsuky that developed a brand new toolset at the very beginning of the year, or activity that falls under the always difficult-to-attribute WinNTI ‘umbrella’. However, and most notably, we can highlight groups such as DarkHotel, LuckyMouse, or even APT10.

The OceanSalt campaign was attributed to APT10, though it’s not very clear how strong the connection is. It seems unlikely that this actor, after the public disclosure and so many years of no known activity, would return with anything that might be attributable to them. At the moment, this is difficult to assess.

LuckyMouse, the second Chinese-speaking group from this list, was very active all year. It hacked national data centers to deploy watering-hole attacks against high-profile victims in central Asia, used a driver signed by a Chinese security-related software developer, and is even suspected of being behind attacks against Oman immediately after the signing of a military agreement with India.

Scarcruft used a new backdoor we call PoorWeb, deployed a 0-day in their campaign at the beginning of the year and used Android malware specially designed for Samsung devices. DarkHotel was also back with a 0-day and new activity, targeting their traditional victims. We were able to establish a connection with a medium level of certainty between DarkHotel and the Konni/Nokki set of activity described by other vendors.

APT10 was especially active against Japanese victims, with new iterations of its malware, as was OceanLotus, which actively deployed watering holes targeting high-profile victims in South Asia with a new custom stager.

In the Middle East we observed groups such as Prince of Persia re-emerge with some activity, along with OilRig. We also detected new MuddyWaters activity, as well as GazaTeam, DesertFalcons and StrongPity among others deploying various campaigns in the region.

On the new kids

At the same time many new sets of activity emerged during the year that were also focused primarily on the Middle East and South East Asia.

This activity was driven by Asian actors such as ShaggyPanther, Sidewinder, CardinalLizard, TropicTrooper, DroppingElephant, Rancor, Tick group, NineBlog, Flyfox and CactusPete – all of them active in the region throughout the year. As a rule, these groups are not that technically advanced, using a variety of approaches to achieve their objectives. They are usually interested in regional targets, with their main objectives being governmental and also military.

In the Middle East we saw activity by LazyMerkaats, FruityArmor, OpParliament, DarkHydrus and DomesticKitten among others. Sets of activity such as that by the Gorgon group are a bit of an exception as they also target victims outside the region.

Finally, we also detected new sets of activity that show an apparent interest in eastern European countries and former Soviet republics. In this group we find DustSquad, ParkingBear and Gallmaker. The latter seems to be interested in overseas embassies as well as military and defense targets in the Middle East.

On the big fishes

Even if some of the activity previously described doesn’t seem that technically advanced, it doesn’t mean it isn’t effective. Looking back we can cite a few public cases where it looks like these attacks are returning to the days when attackers were after major strategic research or blueprints that might be of the interest to state-sponsored groups, and not just some random data.

We have several examples. For instance, APT15 was suspected of targeting a company providing services to military and technology departments of the UK government. Intezer provided extra details about the activity of this group, though it is not clear who the ultimate victim was.

TEMP.Periscope was suspected of hacking maritime organizations related to the South China Sea. It wasn’t the only case in which the industry was targeted, as later it was discovered an unknown actor attacked companies related to Italian naval and defense industries.

Groups such as Thrip showed a clear interest in targeting satellite communication companies and defense organizations in the US and South East Asia.

Finally, the US Naval Undersea Warfare Center was attacked, according to the Washington Post, by a group linked to the Chinese Ministry of State Security, resulting in the theft of 614GB of data and blueprints.

The re-emergence of some of these groups and their victims don’t seem to be a coincidence. Some observers might even see the return of these big targeted attacks as the end of some sort of tacit agreement.

We also observed several attacks against journalists, activists, political dissidents and NGOs around the world. Many of these attacks involved malware developed by companies that provide surveillance tools to governments.

For instance, NSO and its Pegasus malware was discovered in more than 43 countries according to an external investigation, showing that business in this field is blooming. On a darker note, there were reports on how Saudi dissidents and Amnesty International volunteers were targeted with this malware.

The Tibetan community was also specifically targeted with different malware families, including a Linux backdoor, PowerShell payloads, and fake social media to steal credentials.

Finally, CitizenLab provided details of a campaign where Sandvine and GammaGroup artifacts were used for surveillance through local ISPs in Egypt, Turkey and Syria.

On naming and shaming

This is clearly a new strategy, adopted as a defense mechanism and as a response to the attackers, in some cases being justice able to claim individual working for APT groups. This can later be used in diplomatic offensives and lead to tougher consequences at the state level. It seems that governments are no longer shy of making these attacks public and providing details of their investigations, while pointing fingers at the suspected attackers. This is an interesting development and we will see how it evolves in the future.

The end of the Obama-era cyber-agreement between the US and China could be the reason for the wave of Chinese-speaking groups making a comeback, as well as the targeting of some of the high-profile ‘big fishes’ described above. We saw how in this new period of hostility between the two countries, the US obtained the extradition from Belgium of a Chinese intelligence officer charged with conspiring and attempting to commit economic espionage and steal trade secrets from multiple US aviation and aerospace companies.

The US also provided details about a North Korean citizen suspected of being part of the Lazarus group that was behind the Sony Entertainment attack and WannaCry activity, and who is now wanted by the FBI. Maybe in an unrelated note, the US Cert was very active during the year in providing indicators of compromise and detailing Lazarus (HiddenCobra) activity and the tools used by this actor.

After the infamous DNC hack, the US indicted 12 Russian citizens belonging to units 26165 and 74455 of the Russian Main Intelligence Directorate. Seven officers of GRU were also indicted for their alleged role in a campaign to retaliate against the World Anti-Doping Agency that exposed the Russian state-sponsored doping program.

In Europe, UK Officials and the UK National Cyber Security Center attributed the not-Petya attack that took place in June 2017 to Russian military units.

Finally, and in a very interesting initiative, the US Cyber Command launched an ‘information warfare’ campaign with a message to Russian operatives not to even try influencing the US mid-term election process.

All the above, and several other cases, shows how there seems to be a new doctrine in dealing with such hacking attempts, making them public and providing tools for media campaigns, future negotiations and diplomacy, as well as directly targeting operatives.

On hardware

The closer malware gets to the hardware level, the more difficult it is to detect and delete. This is no easy task for the attackers, as it’s usually difficult to find the exploit chain to get that deep in the system, along with the difficulty in developing reliable malware working in such deep levels. That always raises the question of whether this malware already exists, quietly abusing modern CPU architecture characteristics, and we simply don’t see it.

Recent discoveries of vulnerabilities in different processors open the door to exploits that might be around for years, because replacing the CPU is not something that can be easily done. It is not clear yet how Meltdown/Specter and AMDFlaws among others might be exploited and abused in the future, but attackers don’t really need to rush as these vulnerabilities will probably be around for a long time. Even if we haven’t see them being exploited in the wild yet, we believe this is a very valuable piece of knowledge for attackers and maybe also a timely reminder for us all about how important hardware security is.

That leads on to something we actually saw in the VPNFilter attack, in this case targeting networking devices on a massive scale. This campaign, attributed to a Russian-speaking set of activity, allowed attackers to infect hundreds of thousands of devices, providing control of the network traffic as well as allowing MITM attacks. We saw APT actors abusing network devices in the past but never in such an aggressive way.

On other stuff

Triton/Trisis is an industrial-targeting set of activity that gained popularity during the year as it was discovered in some victims, and is suspected of shutting down an oil refinery in an attack where the actor used a 0-day. According to FireEye, this actor might have Russian origins.

In our predictions we already discussed the possibility of destructive attacks becoming normal in situations where tensions exist between two adversaries, using collateral victims to cause harm and send messages in this dangerous grey zone between an open attack and diplomacy.

Financial attackers may not be using very new techniques, but that may be because they don’t need to. The Carbanak group was ‘beheaded’ with the arrest in Spain of one of their leaders; however, that doesn’t seem to have had any impact on subsequent Fin7 activity during the year. They deployed their new Griffon JavaScript backdoor targeting restaurant chains. Meanwhile, a suspected subset of this group – the CobaltGoblin group – was also very active targeting banks in a more direct way.



Securelist

Securelist: KoffeyMaker: notebook vs. ATM

Despite CCTV and the risk of being caught by security staff, attacks on ATMs using a direct connection — so-called black box attacks — are still popular with cybercriminals. The main reason is the low “entry requirements” for would-be cyber-robbers: specialized sites offer both the necessary tools and how-to instructions.

Kaspersky Lab’ experts investigated one such toolkit, dubbed KoffeyMaker, in 2017-2018, when a number of Eastern European banks turned to us for assistance after their ATMs were quickly and almost freely raided. It soon became clear that we were dealing with a black box attack — a cybercriminal opened the ATM, connected a laptop to the cash dispenser, closed the ATM, and left the crime scene, leaving the device inside. Further investigation revealed the “crime instrument” to be a laptop with ATM dispenser drivers and a patched KDIAG tool; remote access was provided through a connection to a USB GPRS modem. The operating system was Windows, most likely XP, ME, or 7 for better driver compatibility.

ATM dispenser connected to a computer without the necessary drivers

ATM dispenser connected to a computer without the necessary drivers

The situation then unfolded according to the usual scenario: the cybercriminal returned at the appointed hour and pretended to use the ATM, while an accomplice remotely connected to the hidden laptop, ran the KDIAG tool, and instructed the dispenser to issue banknotes. The attacker took the money and later retrieved the laptop, too. The whole operation could well be done solo, but the scheme whereby a “mule” handles the cash and ATM side, while a second “jackpotter” provides technical support for a share of the loot, is more common. A single ATM can spit out tens of thousands of dollars, and only hardware encryption between an ATM PC and its dispenser can prevent an attack from occurring.

Overall, the attack was reminiscent of Cutlet Maker, which we described last year, except for the software tools. We were able to reproduce all the steps of KoffeyMaker in our test lab. All the required software was found without too much difficulty. Legitimate tools were used to carry out the attack with the exception of the patched KDIAG utility, which Kaspersky Lab products detect as RiskTool.Win32.DIAGK.a. Note that the same version of this program was previously used by cybercriminals from the Carbanak group.

Hash sums

KDIAG, incl. patched files
49c708aad19596cca380fd02ab036eb2
9a587ac619f0184bad123164f2aa97ca
2e90763ac4413eb815c45ee044e13a43
b60e43d869b8d2a0071f8a2c0ce371aa
3d1da9b83fe5ef07017cf2b97ddc76f1
45d4f8b3ed5a41f830f2d3ace3c2b031
f2c434120bec3fb47adce00027c2b35e
8fc365663541241ad626183d6a48882a
6677722da6a071499e2308a121b9051d
a731270f952f654b9c31850e9543f4ad
b925ce410a89c6d0379dc56c85d9daf0
d7b647f5bcd459eb395e8c4a09353f0d
0bcb612e6c705f8ba0a9527598bbf3f3
ae962a624866391a4321c21656737dcb
83ac7fdba166519b29bb2a2a3ab480f8

Drivers
84c29dfad3f667502414e50a9446ed3f
46972ca1a08cfa1506d760e085c71c20
ff3e0881aa352351e405978e066d9796
4ea7a6ca093a9118df931ad7492cfed5
a8da5b44f926c7f7d11f566967a73a32
f046dc9e38024ab15a4de1bbfe830701
9a1a781fed629d1d0444a3ae3b6e2882

YARA rule

rule software_zz_patched_KDIAG
{
meta:
 author = "Kaspersky Lab"
 filetype = "PE"
 date = "2018-04-28"
 version = "1.0"
 hash = "49c708aad19596cca380fd02ab036eb2"

strings:
$b0 = { 25 80 00 00 00 EB 13 FF 75 EC }
$b1 = { EB 1F 8D 85 FC FE FF FF 50 68 7B 2F 00 00 }
$s0 = "@$MOD$ 040908 0242/0000 CRS1.EXE W32 Copyright (c) Wincor Nixdorf"
condition:
 (
  uint16(0) == 0x5A4D and
  all of ( $s* ) and
  all of ( $b* )
 )
}


Securelist

Securelist: Kaspersky Security Bulletin 2018. Statistics

All the statistics used in this report were obtained using Kaspersky Security Network (KSN), a distributed antivirus network that works with various anti-malware protection components. The data was collected from KSN users who agreed to provide it. Millions of Kaspersky Lab product users from 213 countries and territories worldwide participate in this global exchange of information about malicious activity. All the statistics were collected from November 2017 to October 2018.

The year in figures

  • 30 .01% of user computers were subjected to at least one Malware-class web attack over the year.
  • Kaspersky Lab solutions repelled 1 876 998 691 attacks launched from online resources located all over the world.
  • 554 159 621 unique URLs were recognized as malicious by web antivirus components.
  • Kaspersky Lab’s web antivirus detected 21 643 946 unique malicious objects.
  • 765 538 computers of unique users were targeted by encryptors.
  • 5 638 828 computers of unique users were targeted by miners.
  • Kaspersky Lab solutions blocked attempts to launch malware capable of stealing money via online banking on 830 135 devices.

Fill the form below to download the Kaspersky Security Bulletin 2018. Statistics full report (English, PDF):



Securelist

Securelist: Kaspersky Security Bulletin 2018. Story of the year: miners

Cryptocurrency miners that infect the computers of unsuspecting users essentially operate according to the same business model as ransomware programs: the victim’s computing power is harnessed to enrich the cybercriminals. Only in the case of miners, it might be quite a while before the user notices that 70–80% of their CPU or graphics card power is being used to generate virtual coins. Encrypted documents and ransomware messages are far harder to miss.

Cryptominers usually find their way onto user computers and corporate machines along with adware, hacked games, and other pirated content. What’s more, the present “entry threshold” — that is, the actual process of creating a miner — is rather low: cybercriminals are assisted by ready-to-use affiliate programs, open mining pools, and miner builders. If that weren’t enough, there is another way to steal computing resources through a webpage-embedded mining script that starts when the user opens the site in a browser.  A separate category of cybercriminals are those who target not private computers, but the servers of large companies, for which the infection process is considerably more resource-intense.

2018 began with a rise in the number of miner-related attacks. However, after a drop in the value of the main cryptocurrencies, which lasted from January to February, infection activity noticeably declined. General interest in cryptocurrencies also waned.  Yet the graph clearly shows that while the number of cryptominer attacks decreased, the threat is still current. As for how the November collapse in the Bitcoin exchange rate will affect the number of infections, time will tell.

&&

Number of unique users attacked by miners in Q1–Q3 2018 (download)

Hidden mining software was very popular among botnet owners, as confirmed by our statistics on files downloaded by zombie networks: Q1 2018 saw a boom in cryptominers, and the share of this malware in the first half of the year was 4.6% of the total number of files downloaded by botnets. For comparison, in Q2 2017 this figure was 2.9%. It follows from the data that cybercriminals have come to view botnets as a means of spreading software for mining cryptocurrencies.

H2 2017 H1 2017
1 Lethic 17.0% njRAT 5.2%
2 Neutrino.POS 4.6% Lethic 5.0%
3 njRAT 3.7% Khalesi 4.9%
4 Emotet 3.5% Miners 4.6%
5 Miners 2.9% Neutrino.POS 2.2%
6 Smoke 1.8% Edur 1.3%
7 Cutwail 0.7% PassView 1.3%
8 Ransomware 0.7% Jimmy 1.1%
9 SpyEye 0.5% Gandcrab 1.1%
10 Snojan 0.3% Cutwail 1.1%

Most downloaded threats, H2 2017–H1 2018

Still on the topic of botnets, it is impossible not to mention that in Q3 2018 we registered a decline in the number of DDoS attacks, the most likely reason being, according to our experts, the “reprofiling” of botnets from DDoS attacks to cryptocurrency mining. This was induced not only by the high popularity of cryptocurrencies, but also the high competition in the “DDoS market”, which made the attacks less expensive for clients, but not for the botnetters themselves, who still have to cope with more than a few less-than-legal “organizational issues.”

Mining differs favorably for cybercriminals in that, if executed properly, it can be impossible for the owner of an infected machine to detect, and thus the chances of encountering the cyberpolice are far lower. And the reprofiling of existing server capacity completely hides its owner from the eyes of the law. Evidence suggests that the owners of many well-known botnets have switched their attack vector toward mining.  For example, the DDoS activity of the Yoyo botnet dropped dramatically, although there is no data about it being dismantled.

Moreover, mining has started to command as much (or more) attention as ransomware: this year we encountered several examples of reprofiled malware with added functionality for cryptocurrency mining. And the techniques used by the creators of miners have become more sophisticated.

For instance, an interesting miner implementation, which we dubbed PowerGhost, caught our eye in July this year. The malware can stealthily establish itself in the system and spread inside large corporate networks, infecting workstations and servers alike. To go unnoticed by users and security solutions for as long as possible, the miner employs various fileless techniques. Infection occurs remotely using exploits or remote management tools (Windows Management Instrumentation), and involves running a single-line powershell script that downloads the main body of the malware and immediately starts it without writing to the hard drive.

Another example of reprofiling is the ransomware Trojan Trojan-Ransom.Win32.Rakhni, the first samples of which were detected by Kaspersky Lab back in 2013. Its mining functions are a 2018 innovation. At the same time, their activation depends on whether the folder %AppData%\Bitcoin is present on the infected machine. If it exists, the loader downloads the ransomware. If there is no such folder and, in addition, the computer has more than two logical processors, a miner is downloaded. To keep the malware hidden in the system, the developers made it look like an Adobe product. This can be seen by the icon and the name of the executable file, as well as the fake digital signature, which uses Adobe Systems Incorporated as the company name.

Another piece of malware that has learned how to seed computers with mining utilities is the previously adware-only PBot. The malware spreads through affiliate sites that inject scripts into their pages for redirecting users to sponsored links. The standard distribution scheme looks as follows:

  1. The user visits one of the sites in the affiliate network.
  2. Clicking anywhere on the page causes a new browser window to appear, where an intermediate link opens.
  3. The link directs the user to the PBot download page, which is tasked with downloading and running the malware by deceptive means.

The most common coin among all illegally mined cryptocurrencies is Monero (xmr). This is due to its anonymous algorithm, relatively high market value, and ease of sale, since it is accepted by most major cryptocurrency exchanges. For botnets mining this coin illegally, it is important that CPU resources can be utilized. By some accounts, a total of $175 million has been mined illegally, representing around 5% of all Monero currently in circulation.

Factors affecting the distribution of miners

The conclusion based on data we obtained from various sources is that legislative control over cryptocurrencies has little impact on the spread of hidden mining. For example, in Algeria and Vietnam cryptocurrencies are either prohibited or severely restricted under domestic law. Yet Vietnam is third in the ranking of leading countries by number of miner attacks, and Algeria is sixth. Meanwhile, Iran, which is presently drafting legislation to govern cryptocurrency and developing plans to issue its own “coins,” is in seventh place.

Country Cryptocurrency status % of attacks
Kazakhstan Not prohibited, Not legalized 16.75%
Vietnam Issuance (mining) prohibited 13.00%
Indonesia Recognized as an exchange commodity 12.87%
Ukraine Circulation governed by law 11.19%
Russia Legislation under consideration 10.71%
Algeria Prohibited 9.03%
Iran Legislation in preparation, creation of own cryptocurrency planned 7.21%
India Ban under consideration, hearings in progress 7.20%
Thailand Circulation governed by law 6.76%
Taiwan Not prohibited 5.81%

Top 10 countries by share of miner attacks, January–October 2018 (includes only countries with more than 500,000 Kaspersky Lab clients)

At the other end of the scale, US users were the least affected by cryptominters (1.33% of the total number of attacks), followed by users in Switzerland (1.56%) and Britain (1.66%).

&&

Map representing countries with the lowest share of miner attacks, January–October 2018 (includes only countries with more than 500,000 Kaspersky Lab clients) (download)

The prevalence of miners is not impacted by the cost of electricity, which varies greatly from country to country. Again, this factor is not a consideration for cybercriminals as they exploit third-party resources.

Distribution methods

Looking at the distribution of pirated software in countries with the highest number of miner attacks, one sees a clear correlation: the more freely unlicensed software is distributed, the more miners there are. This is confirmed by our statistics, which indicates that miners most often land on victim computers together with pirated software.

Another penetration vector for miners is adware installers distributed using social engineering. More sophisticated options (for example, propagation through vulnerabilities such as EternalBlue) are aimed at server capacities and are less frequently encountered.

And it should not be forgotten that USB drives have been used to distribute cryptocurrency mining software since at least 2015. The percentage of detections of the popular Bitcoin miner Trojan.Win64.Miner.all on removable devices is growing annually by about one-sixth. In 2018, one in ten users affected by malware transmitted through flash drives was the victim of this particular miner (roughly 9.22%; for comparison, in 2017 it was 6.7%, and in 2016 4.2%).

&&

Millions of unique users found to have malware in the root directory, which is the main sign of infection via removable drives, 2013–2018. Source: KSN (download)

Trojan.Win32.Miner.ays/Trojan.Win.64.Miner.all was detected in India (23.7%), Russia (18.45%), and Kazakhstan (14.38%), but some cases were also logged in Asia, Africa, and Europe (Britain, Germany, the Netherlands, Switzerland, Spain, Belgium, Austria, Italy, Denmark, Sweden), as well as the US, Canada, and Japan.

&&

Share of users impacted by Bitcoin miners on removable drives, 2018. Source: KSN (includes only countries with more than 10,000 Kaspersky Lab clients) (download)

Conclusion

Summing up the past year, we can highlight the following bullet points:

  1. Given the growing value and popularity of cryptocurrencies, cybercriminals are investing resources in the development of new mining technologies, which, according to our data, are gradually replacing ransomware Trojans.
  2. Hidden mining activity declines when cryptocurrency prices fall.
  3. The spread of hidden mining is not impacted by factors such as domestic legislative control or cost of electricity.
  4. Miners often get on victims’ computers during the download of unlicensed content or installation of pirated software. As a consequence, this type of threat is most prevalent in countries with poor regulation of the unlicensed software market, as well a low level of overall digital literacy among users.

Kaspersky Security Bulletin 2018. Story of the year: miners” (English, PDF)



Securelist

Securelist: Threat predictions for industrial security in 2019

The past few years have been very intense and eventful when it comes to incidents affecting the information security of industrial systems. That includes new vulnerabilities, new threat vectors, accidental infections of industrial systems and detected targeted attacks. In response, last year we developed some Threat Predictions for Industrial Security in 2018, outlining the trends most likely to unfold in the year ahead.

The industrial cybersecurity threat landscape moves at a slower and more rigid pace than the information technology threat landscape in general. Attacks on ICS are still hard to monetize. Industrial organizations are still out of scope for the majority of cybercriminals. They are a relatively new target for adversaries who have already started attacking them. These are still applying existing tools and tactics to their attacks. That is why the majority of the industrial threat predictions from last year are still unfolding, although some of them have already come true.

Kaspersky Lab specialists have spent a few years investigating the cyberthreat landscape for industrial organizations and trying to bring their expertise and technology to OT environments. We are still on a long journey, with various to difficulties cope with and problems yet to solve. Constantly keeping in contact with many researchers in other security organizations and some ICS security pioneers from inside industrial companies; we have come to the conclusion that some of the difficulties we face are common to the industry. Solving some of those is mandatory to make the world more secure and safe.

So, although the fog of 2018’s predictions and threat landscape has yet to clear, we decided to focus on the major problems likely to affect the work of professionals involved in industrial systems in 2019.

Top four cybersecurity challenges facing industrial enterprises in 2019

The ever-increasing attack surface

The increasing amount of automation systems, the variety of automation tools, number of organizations and individuals with direct or remote access to automation systems, as well as the emergence of communication channels for monitoring and remote control between previously independent objects – all expand the opportunities for criminals to plan and execute their attacks.

Growing interest of cybercriminals and special services

A decrease in profitability and increase in risks from cyberattacks aimed at traditional victims is pushing criminals to search for new targets, including those within industrial organizations.

At the same time, special services in many countries, as well as other organized groups – motivated by internal and external political interests – and financially-motivated groups, are actively engaged in the research and development of techniques to implement espionage and terrorist attacks aimed at industrial enterprises.

Taking into account the current geopolitical context, the development of industrial enterprises’ automation systems, and the transition to new management processes and models of production and economic activity, this situation will continue to develop in the coming years, negatively affecting industrial organizations.

The underestimation of general threat levels

A lack of public access to information about information security issues within industrial enterprises, coupled with the relative rarity of targeted attacks on automation systems, an excessive belief in emergency protection systems and the denial of objective reality is having a negative effect on the assessment of threat levels by owners and operators of industrial enterprises and their personnel.

The misunderstanding of threat specifics and the suboptimal choice of protection options

In the world of industrial cybersecurity, several high–profile incidents carried out with the help of targeted attacks against a very limited number of victims, created an information landscape that formed fully the idea of a potential threat – both among information security researchers and security developers, and among potential users of these tools.

However, the professional reporting of these incidents was often too difficult to understand by the majority of potential users, and was devoid of important OT details. The information field formed in these conditions, including the absence of a daily need to deflect the attacks aimed at automated control systems, gave developers a chance to create products that might protect better from the artificial scenarios thought up by researchers themselves, than from real world day-to-day threats. This could leave the automation systems of industrial enterprises vulnerable to real life attacks, including random ones and targeted attack campaigns organized by cyber criminals.

Full version of the threat predictions will be published on ICS CERT website.

Full report “Kaspersky Security Bulletin: Threat predictions for industrial security in 2019″ (English, PDF)



Securelist

Securelist: Cryptocurrency threat predictions for 2019

Introduction – key events in 2018

2018 saw cryptocurrency become an established part of many people’s lives, and a more attractive target for cybercriminals across the world. To some extent, the malicious mining of cryptocurrencies even prevailed over the main threat of the last few years: ransomware.

However, in the second half of 2018, the blockchain and cryptocurrency industry faced a major development: falling prices for cryptocurrencies. The impact was felt across the landscape, with rapid decline in public interest, the activity of the crypto community and traders, and in the related activity of cybercriminals.

While this will certainly affect our forecasts for 2019, let’s see how the forecasts we made for this year worked out.

1. ‘Ransomware attacks will force users to buy cryptocurrency’

This prediction turned out to be partially true. In 2018, we saw a decline in the popularity of encryptors, combined with a rise in the malicious use of cryptocurrency miners. It transpired that it is safer for attackers to perform discreet mining on infected devices than to demand a ransom and attract attention. However, it is too early to dismiss ransomware as a major threat; it is still an effective method of infection and monetization of both individuals and organizations – and cryptocurrencies remain a more easily anonymized form of ransom payment.

2. ‘We will see targeted attacks with malicious miners’

This prediction did not come true. We observed mainly isolated incidents where miners were maliciously installed in an infected corporate network. There are several reasons for that:

  • Companies have learned to detect miners that are run on the computers of employees/administrators; both those installed by users themselves and by third parties without the knowledge of the user.
  • The attackers themselves do not appear to consider this a promising approach. Targeted and sophisticated attacks are more about gaining persistence in the network for the purpose of espionage or the theft of money or data. It is therefore better not to attract attention by crypto-mining.

3. ‘The rise of miners will continue and involve new actors’

This prediction also turned out to be partially true: the malicious use of cryptocurrency miners actively increased during the first quarter of 2018, peaking in March. Over the following months there was a gradual decrease in activity due to the drop in price for cryptocurrencies.

4. ‘There will be more web-mining’

Again, this prediction turned out to be partially true. The web mining of cryptocurrencies reached a peak in January 2018, after which it began to decline. Webmasters, hoping to use web mining as an alternative means of website monetization alongside advertising, did not usually notify users about any hidden mining taking place on their sites. This meant that web mining quickly became associated with malicious activity. After that, it was difficult to restore its reputation.

5. ‘The fall of ICOs (Initial Coin Offering)’

Yes and no. On the one hand, collecting money with the help of ICOs continued: projects became larger and the fees did not fall. On the other hand, many projects that collected impressive amounts through ICOs in 2017 were not be able to create the promised product in time during 2018, which inevitably affected the exchange price of the sold tokens.

Top three predictions for 2019

1. Excessive expectations about the use of blockchain beyond the cryptocurrency sphere will disappear

In the end, we expect this trend to be driven by people rather than the technology’s capability, as organizations and industries come to the conclusion that blockchain has a rather narrow scope of application, and most attempts to use in different ways are not justified. The reliable application of blockchain beyond cryptocurrency has been explored and experimented with for years, but there is little evidence of achievement. We expect 2019 to be the year people stop trying.

2. Cryptocurrencies as a means of payment will decline further

In 2017 a number of suppliers of goods and services announced that they would accept cryptocurrencies as a form of payment. However, in the face of huge commissions (an acute problem in December 2017), slow transfers, a large price for integration, and, most importantly, a small number of customers, its use as a method of payment declined steadily. In the end, the use of cryptocurrencies by a legitimate business simply does not make much sense.

3. There will be no return to 2017’s sky-high exchange rates

Until January 2018, there were immense highs and lows in the price of Bitcoin. But we do not expect these to return as the value of cryptocurrencies levels out to reflect their popularity. We believe there is a finite audience for whom cryptocurrencies are of interest, and once that limit is reached the price will not rise further.

 “Cryptocurrency threat predictions for 2019” (PDF)



Securelist

Securelist: The Rotexy mobile Trojan – banker and ransomware

On the back of a surge in Trojan activity, we decided to carry out an in-depth analysis and track the evolution of some other popular malware families besides Asacub. One of the most interesting and active specimens to date was a mobile Trojan from the Rotexy family. In a three-month period from August to October 2018, it launched over 70,000 attacks against users located primarily in Russia.

An interesting feature of this family of banking Trojans is the simultaneous use of three command sources:

  • Google Cloud Messaging (GCM) service – used to send small messages in JSON format to a mobile device via Google servers;
  • malicious C&C server;
  • incoming SMS messages.

This ‘versatility’ was present in the first version of Rotexy and has been a feature of all the family’s subsequent representatives. During our research we also arrived at the conclusion that this Trojan evolved from an SMS spyware Trojan that was first spotted in October 2014. Back then it was detected as Trojan-Spy.AndroidOS.SmsThief, but later versions were assigned to another family ­– Trojan-Banker.AndroidOS.Rotexy.

The modern version of Rotexy combines the functions of a banking Trojan and ransomware. It spreads under the name AvitoPay.apk (or similar) and downloads from websites with names like youla9d6h.tk, prodam8n9.tk, prodamfkz.ml, avitoe0ys.tk, etc. These website names are generated according to a clear algorithm: the first few letters are suggestive of popular classified ad services, followed by a random string of characters, followed by a two-letter top-level domain. But before we go into the details of what the latest version of Rotexy can do and why it’s distinctive, we would like to give a summary of the path the Trojan has taken since 2014 up to the present day.

Evolution of Rotexy

2014–2015

Since the malicious program was detected in 2014, its main functions and propagation method have not changed: Rotexy spreads via links sent in phishing SMSs that prompt the user to install an app. As it launches, it requests device administrator rights, and then starts communicating with its C&C server.

A typical class list in the Trojan’s DEX file

Until mid-2015, Rotexy used a plain-text JSON format to communicate with its C&C. The C&C address was specified in the code and was also unencrypted:

In some versions, a dynamically generated low-level domain was used as an address:

In its first communication, the Trojan sent the infected device’s IMEI to the C&C, and in return it received a set of rules for processing incoming SMSs (phone numbers, keywords and regular expressions) – these applied mainly to messages from banks, payment systems and mobile network operators. For instance, the Trojan could automatically reply to an SMS and immediately delete it.

Message to C&C requesting an SMS processing template, and the server’s reply

Rotexy then sent information about the smartphone to the C&C, including the phone model, number, name of the mobile network operator, versions of the operating system and IMEI.

With each subsequent request, a new subdomain was generated. The algorithm for generating the lowest-level domain name was hardwired in the Trojan’s code.

The Trojan also registered in Google Cloud Messaging (GCM), meaning it could then receive commands via that service. The Trojan’s list of possible commands has remained practically unchanged throughout its life, and will be described below in detail.

The Trojan’s assets folder contained the file data.db with a list of possible values for the User-Agent field for the PAGE command (which downloads the specified webpage). If the value of this field failed to arrive from the C&C, it was selected from the file data.db using a pseudo-random algorithm.

Contents of data.db

2015–2016

Starting from mid-2015, the Trojan began using the AES algorithm to encrypt data communicated between the infected device and the C&C:

Also starting with the same version, data is sent in a POST request to the relative address with the format “/[number]” (a pseudo-randomly generated number in the range 0–9999).

In some samples, starting from January 2016, an algorithm has been implemented for unpacking the encrypted executable DEX file from the assets folder. In this version of Rotexy, dynamic generation of lowest-level domains was not used.

2016

From mid-2016 on, the cybercriminals returned to dynamic generation of lowest-level domains. No other significant changes were observed in the Trojan’s network behavior.

Query from the Trojan to the C&C

In late 2016, versions of the Trojan emerged that contained the card.html phishing page in the assets/www folder. The page was designed to steal users’ bank card details:

2017–2018

From early 2017, the HTML phishing pages bank.html, update.html and extortionist.html started appearing in the assets folder. Also, in some versions of the Trojan the file names were random strings of characters.

In 2018, versions of Rotexy emerged that contacted the C&C using its IP address. ‘One-time’ domains also appeared with names made up of random strings of characters and numbers, combined with the top-level domains .cf, .ga, .gq, .ml, or .tk.

At this time, the Trojan also began actively using different methods of obfuscation. For example, the DEX file is packed with garbage strings and/or operations, and contains a key to decipher the main executable file from the APK.

Latest version (2018)

Let’s now return to the present day and a detailed description of the functionality of a current representative of the Rotexy family (SHA256: ba4beb97f5d4ba33162f769f43ec8e7d1ae501acdade792a4a577cd6449e1a84).

Application launch

When launching for the first time, the Trojan checks if it is being launched in an emulation environment, and in which country it is being launched. If the device is located outside Russia or is an emulator, the application displays a stub page:

In this case, the Trojan’s logs contain records in Russian with grammatical errors and spelling mistakes:

If the check is successful, Rotexy registers with GCM and launches SuperService which tracks if the Trojan has device administrator privileges. SuperService also tracks its own status and relaunches if stopped. It performs a privilege check once every second; if unavailable, the Trojan starts requesting them from the user in an infinite loop:

If the user agrees and gives the application the requested privileges, another stub page is displayed, and the app hides its icon:

If the Trojan detects an attempt to revoke its administrator privileges, it starts periodically switching off the phone screen, trying to stop the user actions. If the privileges are revoked successfully, the Trojan relaunches the cycle of requesting administrator privileges.

If, for some reason, SuperService does not switch off the screen when there is an attempt to revoke the device administrator privileges, the Trojan tries to intimidate the user:

While running, Rotexy tracks the following:

  • switching on and rebooting of the phone;
  • termination of its operation – in this case, it relaunches;
  • sending of an SMS by the app – in this case, the phone is switched to silent mode.

C&C communications

The default C&C address is hardwired in the Rotexy code:

The relative address to which the Trojan will send information from the device is generated in a pseudo-random manner. Depending on the Trojan version, dynamically generated subdomains can also be used.

In this sample of the Trojan, the Plugs.DynamicSubDomain value is false, so subdomains are not generated

The Trojan stores information about C&C servers and the data harvested from the infected device in a local SQLite database.

First off, the Trojan registers in the administration panel and receives the information it needs to operate from the C&C (the SMS interception templates and the text that will be displayed on HTML pages):

Rotexy intercepts all incoming SMSs and processes them according to the templates it received from the C&C. Also, when an SMS arrives, the Trojan puts the phone into silent mode and switches off the screen so the user doesn’t notice that a new SMS has arrived. When required, the Trojan sends an SMS to the specified phone number with the information it has received from the intercepted message. (It is specified in the interception template whether a reply must be sent, and which text should be sent to which address.) If the application hasn’t received instructions about the rules for processing incoming SMSs, it simply saves all SMSs to a local database and uploads them to the C&C.

Apart from general information about the device, the Trojan sends a list of all the running processes and installed applications to the C&C. It’s possible the threat actors use this list to find running antivirus or banking applications.

Rotexy will perform further actions after it receives the corresponding commands:

  • START, STOP, RESTART — start, stop, restart SuperService.
  • URL — update C&C address.
  • MESSAGE – send SMS containing specified text to a specified number.
  • UPDATE_PATTERNS – reregister in the administration panel.
  • UNBLOCK – unblock the telephone (revoke device administrator privileges from the app).
  • UPDATE – download APK file from C&C and install it. This command can be used not just to update the app but to install any other software on the infected device.
  • CONTACTS – send text received from C&C to all user contacts. This is most probably how the application spreads.
  • CONTACTS_PRO – request unique message text for contacts from the address book.
  • PAGE – contact URL received from C&C using User-Agent value that was also received from C&C or local database.
  • ALLMSG – send C&C all SMSs received and sent by user, as stored in phone memory.
  • ALLCONTACTS – send all contacts from phone memory to C&C.
  • ONLINE – send information about Trojan’s current status to C&C: whether it has device administrator privileges, which HTML page is currently displayed, whether screen is on or off, etc.
  • NEWMSG – write an SMS to the device memory containing the text and sender number sent from C&C.
  • CHANGE_GCM_ID – change GSM ID.
  • BLOCKER_BANKING_START – display phishing HTML page for entry of bank card details.
  • BLOCKER_EXTORTIONIST_START – display HTML page of the ransomware.
  • BLOCKER_UPDATE_START – display fake HTML page for update.
  • BLOCKER_STOP – block display of all HTML pages.

The C&C role for Rotexy can be filled not only by a web server but also by any device that can send SMSs. The Trojan intercepts incoming SMSs and can receive the following commands from them:

  • “3458” — revoke device administrator privileges from the app;
  • “hi”, “ask” — enable and disable mobile internet;
  • “privet”, “ru” — enable and disable Wi-Fi;
  • “check” — send text “install: [device IMEI]” to phone number from which SMS was sent;
  • “stop_blocker” — stop displaying all blocking HTML pages;
  • “393838” — change C&C address to that specified in the SMS.

Information about all actions performed by Rotexy is logged in the local database and sent to the C&C. The server then sends a reply that contains instructions on further actions to be taken.

Displaying HTML pages

We’ll now look at the HTML pages that Rotexy displays and the actions performed with them.

  • The Trojan displays a fake HTML update page (update.html) that blocks the device’s screen for a long period of time.
  • The Trojan displays the extortion page (extortionist.html) that blocks the device and demands a ransom for unblocking it. The sexually explicit images in this screenshot have been covered with a black box.
  • The Trojan displays a phishing page (bank.html) prompting the user to enter their bank card details. This page mimics a legitimate bank form and blocks the device screen until the user enters all the information. It even has its own virtual keyboard that supposedly protects the victim from keyloggers.

In the areas marked ‘{text}’ Rotexy displays the text it receives from the C&C. Typically, it is a message saying that the user has received a money transfer, and that they must enter their bank card details so the money can be transferred to their account.

The entered data is then checked and the last four digits of the bank card number are also checked against the data sent in the C&C command. The following scenario may play out: according to the templates for processing incoming SMSs, Rotexy intercepts a message from the bank that contains the last four digits of the bank card connected to the phone number. The Trojan sends these digits to the C&C, which in turn sends a command to display a fake data entry window to check the four digits. If the user has provided the details of another card, then the following window is displayed:

Screenshot displaying the message: “You have entered an incorrect card. Enter the card ending in the digits: 1234”

The application leaves the user with almost no option but to enter the correct card number, as it checks the entered number against the bank card details the cybercriminals received earlier.

When all the necessary card details are entered and have been checked, all the information is uploaded to the C&C.

How to unblock the phone

Now for some good news: Rotexy doesn’t have a very well-designed module for processing commands that arrive in SMSs. It means the phone can be unblocked in some cases when it has been blocked by one of the above HTML pages. This is done by sending “3458” in an SMS to the blocked device – this will revoke the administrator privileges from the Trojan. After that it’s necessary to send “stop_blocker” to the same number – this will disable the display of HTML pages that extort money and block the screen. Rotexy may start requesting device administrator privileges again in an infinite loop; in that case, restart the device in safe mode and remove the malicious program.

However, this method may not work if the threat actors react quickly to an attempt to remove the Trojan. In that case, you first need to send the text “393838” in an SMS to the infected device and then repeat all the actions described above; that text message will change the C&C address to “://”, so the phone will no longer receive commands from the real C&C.

Please note that these unblocking instructions are based on an analysis of the current version of Rotexy and have been tested on it. However, it’s possible the set of commands may change in future versions of the Trojan.

Geography of Rotexy attacks

According to our data, 98% of all Rotexy attacks target users in Russia. Indeed, the Trojan explicitly targets Russian-speaking users. There have also been cases of users in Ukraine, Germany, Turkey and several other countries being affected.

Kaspersky Internet Security for Android and the Sberbank Online app securely protect users against attacks by this Trojan.

IOCs

SHA256
0ca09d4fde9e00c0987de44ae2ad51a01b3c4c2c11606fe8308a083805760ee7
4378f3680ff070a1316663880f47eba54510beaeb2d897e7bbb8d6b45de63f96
76c9d8226ce558c87c81236a9b95112b83c7b546863e29b88fec4dba5c720c0b
7cc2d8d43093c3767c7c73dc2b4daeb96f70a7c455299e0c7824b4210edd6386
9b2fd7189395b2f34781b499f5cae10ec86aa7ab373fbdc2a14ec4597d4799ba
ac216d502233ca0fe51ac2bb64cfaf553d906dc19b7da4c023fec39b000bc0d7
b1ccb5618925c8f0dda8d13efe4a1e1a93d1ceed9e26ec4a388229a28d1f8d5b
ba4beb97f5d4ba33162f769f43ec8e7d1ae501acdade792a4a577cd6449e1a84
ba9f4d3f4eba3fa7dce726150fe402e37359a7f36c07f3932a92bd711436f88c
e194268bf682d81fc7dc1e437c53c952ffae55a9d15a1fc020f0219527b7c2ec

С&C

2014–2015:

  • secondby.ru
  • darkclub.net
  • holerole.org
  • googleapis.link

2015–2016:

  • test2016.ru
  • blackstar.pro
  • synchronize.pw
  • lineout.pw
  • sync-weather.pw

2016

  • freedns.website
  • streamout.space

2017–2018:

  • streamout.space
  • sky-sync.pw
  • gms-service.info


Securelist

Securelist: Kaspersky Security Bulletin: Threat Predictions for 2019

There’s nothing more difficult than predicting. So, instead of gazing into a crystal ball, the idea here is to make educated guesses based on what has happened recently and where we see a trend that might be exploited in the coming months.

Asking the most intelligent people I know, and basing our scenario on APT attacks because they traditionally show the most innovation when it comes to breaking security, here are our main ‘predictions’ of what might happen in the next few months.

No more big APTs

What? How is it possible that in a world where we discover more and more actors every day the first prediction seems to point in the opposite direction?

The reasoning behind this is that the security industry has consistently discovered highly sophisticated government-sponsored operations that took years of preparation. What seems to be a logical reaction to that situation from an attacker’s perspective would be exploring new, even more sophisticated techniques that are much more difficult to discover and to attribute to specific actors.

Indeed, there are many different ways of doing this. The only requirement would be an understanding of the techniques used by the industry for attribution and for identifying similarities between different attacks and the artifacts used in them– something that doesn’t seem to be a big secret. With sufficient resources, a simple solution for an attacker could be having different ongoing sets of activity that are very difficult to relate to the same actor or operation. Well-resourced attackers could start new innovative operations while keeping their old ones alive. Of course, there’s still a good chance of the older operations being discovered, but discovering the new operations would pose a greater challenge.

Instead of creating more sophisticated campaigns, in some cases it appears to be more efficient for some very specific actors who have the capability to do so, to directly target infrastructure and companies where victims can be found, such as ISPs. Sometimes this can be accomplished through regulation, without the need for malware.

Some operations are simply externalized to different groups and companies that use different tools and techniques, making attribution extremely difficult. It’s worth keeping in mind that in the case of government-sponsored operations this ‘centrifugation’ of resources and talent might affect the future of such campaigns. Technical capabilities and tools are owned by the private industry in this scenario, and they are for sale for any customer that, in many cases, doesn’t fully understand the technical details and consequences behind them.

All this suggests that we’re unlikely to discover new highly sophisticated operations – well-resourced attackers are more likely to simply shift to new paradigms.

Networking hardware and IOT

It just seemed logical that at some point every actor would deploy capabilities and tools designed to target networking hardware. Campaigns like VPNFilter were a perfect example of how attackers have already started deploying their malware to create a multipurpose ‘botnet’. In this particular case, even when the malware was extremely widespread, it took some time to detect the attack, which is worrisome considering what might happen in more targeted operations.

Actually, this idea can go even further for well-resourced actors: why not directly target even more elemental infrastructure instead of just focusing on a target organization? We haven’t reached that level of compromise (to our knowledge), but it was clear from past examples (like Regin) how tempting that level of control is for any attacker.

Vulnerabilities in networking hardware allow attackers to follow different directions. They might go for a massive botnet-style compromise and use that network in the future for different goals, or they might approach selected targets for more clandestine attacks. In this second group we might consider ‘malware-less’ attacks, where opening a VPN tunnel to mirror or redirect traffic might provide all the necessary information to an attacker.

All these networking elements might also be part of the mighty IoT, where botnets keep growing at an apparently unstoppable pace. These botnets could be incredibly powerful in the wrong hands when it comes to disrupting critical infrastructure, for instance. This can be abused by well-resourced actors, possibly using a cover group, or in some kind of terror attack.

One example of how these versatile botnets can be used, other than for disruptive attacks, is in short-range frequency hopping for malicious communications, avoiding monitoring tools by bypassing conventional exfiltration channels.

Even though this seems to be a recurrent warning year after year, we should never underestimate IoT botnets – they keep growing stronger.

Public retaliation

One of the biggest questions in terms of diplomacy and geopolitics was how to deal with an active cyberattack. The answer is not simple and depends heavily on how bad and blatant the attack was, among many other considerations. However, it seems that after hacks like that on the Democratic National Committee, things became more serious.

Investigations into recent high-profile attacks, such as the Sony Entertainment Network hacks or the attack on the DNC, culminated in a list of suspects being indicted. That results not only in people facing trial but also a public show of who was behind the attack. This can be used to create a wave of opinion that might be part of an argument for more serious diplomatic consequences.

Actually we have seen Russia suffering such consequences as a result of their alleged interference in democratic processes. This might make others rethink future operations of this kind.

However, the fear of something like that happening, or the thought that it might already have happened, was the attackers’ biggest achievement. They can now exploit such fear, uncertainty and doubt in different, more subtle ways – something we saw in notable operations, including that of the Shadowbrokers. We expect more to come.

What will we see in the future? The propaganda waters were probably just being tested by past operations. We believe this has just started and it will be abused in a variety of ways, for instance, in false flag incidents like we saw with Olympic Destroyer, where it’s still not clear what the final objective was and how it might have played out.

Emergence of newcomers

Simplifying somewhat, the APT world seems to be breaking into two groups: the traditional well-resourced most advanced actors (that we predict will vanish) and a group of energetic newcomers who want to get in on the game.

The thing is that the entry barrier has never been so low, with hundreds of very effective tools, re-engineered leaked exploits and frameworks of all kinds publicly available for anyone to use. As an additional advantage, such tools make attribution nearly impossible and can be easily customized if necessary.

There are two regions in the world where such groups are becoming more prevalent: South East Asia and the Middle East. We have observed the rapid progression of groups suspected of being based in these regions, traditionally abusing social engineering for local targets, taking advantage of poorly protected victims and the lack of a security culture. However, as targets increase their defenses, attackers do the same with their offensive capabilities, allowing them to extend their operations to other regions as they improve the technical level of their tools. In this scenario of scripting-based tools we can also find emerging companies providing regional services who, despite OPSEC failures, keep improving their operations.

One interesting aspect worth considering from a more technical angle is how JavaScript post-exploitation tools might find a new lease of life in the short term, given the difficulty of limiting its functionality by an administrator (as opposed to PowerShell), its lack of system logs and its ability to run on older operating systems.

The negative rings

The year of Meltdown/Specter/AMDFlaws and all the associated vulnerabilities (and those to come) made us rethink where the most dangerous malware actually lives. And even though we have seen almost nothing in the wild abusing vulnerabilities below Ring 0, the mere possibility is truly scary as it would be invisible to almost all the security mechanisms we have.

For instance, in the case of SMM there has at least been a publicly available PoC since 2015. SMM is a CPU feature that would effectively provide remote full access to a computer without even allowing Ring 0 processes to have access to its memory space. That makes us wonder whether the fact that we haven’t found any malware abusing this so far is simply because it is so difficult to detect. Abusing this feature seems to be too good an opportunity to ignore, so we are sure that several groups have been trying to exploit such mechanisms for years, maybe successfully.

We see a similar situation with virtualization/hypervisor malware, or with UEFI malware. We have seen PoCs for both, and HackingTeam even revealed a UEFI persistence module that’s been available since at least 2014, but again no real ITW examples as yet.

Will we ever find these kinds of unicorns? Or haven’t they been exploited yet? The latter possibility seems unlikely.

Your favorite infection vector

In probably the least surprising prediction of this article we would like to say a few words about spear phishing. We believe that the most successful infection vector ever will become even more important in the nearest future. The key to its success remains its ability to spark the curiosity of the victim, and recent massive leaks of data from various social media platforms might help attackers improve this approach.

Data obtained from attacks on social media giants such as Facebook and Instagram, as well as LinkedIn and Twitter, is now available on the market for anyone to buy. In some cases, it is still unclear what kind of data was targeted by the attackers, but it might include private messages or even credentials. This is a treasure trove for social engineers, and could result in, for instance, some attacker using the stolen credentials of some close contact of yours to share something on social media that you already discussed privately, dramatically improving the chances of a successful attack.

This can be combined with traditional scouting techniques where attackers double-check the target to make sure the victim is the right one, minimizing the distribution of malware and its detection. In terms of attachments, it is fairly standard to make sure there is human interaction before firing off any malicious activity, thus avoiding automatic detection systems.

Indeed, there are several initiatives using machine learning to improve phishing’s effectiveness. It’s still unknown what the results would be in a real-life scenario, but what seems clear is that the combination of all these factors will keep spear phishing as a very effective infection vector, especially via social media in the months to come.

Destructive destroyer

Olympic destroyer was one of the most famous cases of potentially destructive malware during the past year, but many attackers are incorporating such capabilities in their campaigns on a regular basis. Destructive attacks have several advantages for attackers, especially in terms of creating a diversion and cleaning up any logs or evidence after the attack. Or simply as a nasty surprise for the victim.

Some of these destructive attacks have geostrategic objectives related to ongoing conflicts as we have seen in Ukraine, or with political interests like the attacks that affected several oil companies in Saudi Arabia. In some other cases they might be the result of hacktivism, or activity by a proxy group that’s used by a more powerful entity that prefers to stay in the shadows.

Anyway, the key to all these attacks is that they are ‘too good’ not to use. In terms of retaliation for instance, governments might use them as a response ranged somewhere between a diplomatic answer and an act of war, and indeed some governments are experimenting with them. Most of these attacks are planned in advance, which involves an initial stage of reconnaissance and intrusion. We don’t know how many potential victims are already in this situation where everything is ready, just waiting for the trigger to be pulled, or what else the attackers have in their arsenal waiting for the order to attack.

ICS environments and critical infrastructure are especially vulnerable to such attacks, and even though industry and governments have put a lot of effort in over the last few years to improve the situation, things are far from ideal. That’s why we believe that even though such attacks will never be widespread, in the next year we expect to see some occurring, especially in retaliation to political decisions.

Advanced supply chain

This is one of the most worrisome vectors of attack, which has been successfully exploited over the last two years, and it has made everyone think about how many providers they have and how secure they are. Well, there is no easy answer to this kind of attack.

Even though this is a fantastic vector for targeting a whole industry (similar to watering hole attacks) or even a whole country (as seen with NotPetya), it’s not that good when it comes to more targeted attacks as the risk of detection is higher. We have also seen more indiscriminate attempts like injecting malicious code in public repositories for common libraries. The latter technique might be useful in very carefully timed attacks when these libraries are used in a very particular project, with the subsequent removal of the malicious code from the repository.

Now, can this kind of attack be used in a more targeted way? It appears to be difficult in the case of software because it will leave traces everywhere and the malware is likely to be distributed to several customers. It is more realistic in cases when the provider works exclusively for a specific customer.

What about hardware implants? Are they a real possibility? There has been some recent controversy about that. Even though we saw from Snowden’s leaks how hardware can be manipulated on its way to the customer, this does not appear to be something that most actors can do other than the very powerful ones. And even they will be limited by several factors.

However, in cases where the buyer of a particular order is known, it might be more feasible for an actor to try and manipulate hardware at its origin rather than on its way to the customer.

It’s difficult to imagine how all the technical controls in an industrial assembly line could be circumvented and how such manipulation could be carried out. We don’t want to discard this possibility, but it would probably entail the collaboration of the manufacturer.

All in all, supply chain attacks are an effective infection vector that we will continue to see. In terms of hardware implants we believe it is extremely unlikely to happen and if it does, we will probably never know….

And mobile

This is in every year’s predictions. Nothing groundbreaking is expected, but it’s always interesting to think about the two speeds for this slow wave of infections. It goes without saying that all actors have mobile components in their campaigns; it makes no sense only going for PCs. The reality is that we can find many examples of artifacts for Android, but also a few improvements in terms of attacking iOS.

Even though successful infections for iPhone requires concatenating several 0-days, it’s always worth remembering that incredibly well-resourced actors can pay for such technology and use it in critical attacks. Some private companies claim they can access any iPhone that they physically possess. Other less affluent groups can find some creative ways to circumvent security on such devices using, for instance, rogue MDM servers and asking targets through social engineering to use them in their devices, providing the attackers with the ability to install malicious applications.

It will be interesting to see if the boot code for iOS leaked at the beginning of the year will provide any advantage to the attackers, or if they’ll find new ways of exploiting it.

In any case, we don’t expect any big outbreak when it comes to mobile targeted malware, but we expect to see continuous activity by advanced attackers aimed at finding ways to access their targets’ devices.

The other things

What might attackers be thinking about in more futuristic terms? One of the ideas, especially in the military field, might be to stop using weak error-prone humans and replacing them with something more mechanical. With that in mind, and also thinking of the alleged GRU agents expelled from the Netherlands last April after trying to hack into the OPCW’s Wi-Fi network as an example, what about using drones instead of human agents for short-range hacking?

Or what about backdooring some of the hundreds of cryptocurrency projects for data gathering, or even financial gain?

Use of any digital good for money laundering? What about using in-game purchases and then selling such accounts later in the marketplace?

There are so many possibilities that predictions always fall short of reality. The complexity of the environment cannot be fully understood anymore, raising possibilities for specialist attacks in different areas. How can a stock exchange’s internal inter-banking system be abused for fraud? I have no idea, I don’t even know if such a system exists. This is just one example of how open to the imagination the attackers behind these campaigns are.

We are here to try and anticipate, to understand the attacks we don’t, and to prevent them from occurring in the future.

Full report “Kaspersky Security Bulletin: Threat Predictions for 2019” (English, PDF)



Securelist