The ABC’s of APTs: Shamoon

sham35Welcome to the grey zone where politics and cyber meet. APTs or advanced persistent threats, are one of my favourite acronyms (but then you know how I am intrigued by Stuxnet and cartels), and essentially are how nation states get their digital digs at each other. Usually the intention is to get information, because knowledge is power. Cyberespionage can give a competing nation a real competitive advantage in the world economy, among other things. But sometimes, there is a need to control more, and that is where weaponizing code takes on a whole new nasty.

The keyword here is “persistence.”  First, attackers must find their way into the networks of the target. Usually, they employ targeted spear phishing, painstakingly staking out the right victim to receive that loaded email.  The investment of time and money at this point is essential, so as not to tip anyone off. And the emails are crafted so carefully, picking up on points tailored to that recipient so that they will open it, and launch the attachment that will create an entry point for the attacker. There is a reason why phishing is at the heart of so many breaches.

Now, imagine a video game, where you must progressively meet the challenges of each level to go higher. That is the attacker moving through the network, acquiring credentials to gain access to the crown jewels. The strategy is to find someone lower level, then work your way up. Hence, persistence, because this is an investment of both time and patience. Expect the key executives or decision makers to be well-guarded, with access and authorization controls in place. Not the case for someone lower on the food chain. All an attacker needs is to gain access. As proven repeatedly, once in, they can take all the time they need to find what they want. Case in point: the attack on the Ukraine power grid in December 2016.  The attackers were in that system for over nine months, collecting what they needed, notably credentials for the Virtual Private Network, that enabled them to jump the security gap onto the restricted side. As Stuxnet taught us, there is no such thing as air-gapped security.

shamoonattackgraphic

We know the Russians hacked the US; we know China hacked the US and Canada; and yes, the US has hacked someone too. These are the games nations play. The trick, of course, is not to get caught before you have the prize. And when you do get caught?  Well, as we’ve seen play out, nothing really bad happens. Just expect that your victim will be in your systems. Unless information isn’t the endgame and control is. Then, be prepared for something to go bump in the night.

Shamoon is devastating wiper malware that took out a massive swath of Saudi Aramco when it first debuted in 2012.  Linked to Iran, and an ongoing feud in the region between key players, it was a targeted attack against the oil giant, damaging or destroying 35,000 computers. Sec Def at the time, Leon Panetta, described it as “probably the most destructive cyber attack on a business.”

Wiper malware was used against business targets in  December 2014 destroying the systems in a Vegas casino, The Sands, after owner Sheldon Adelson advocated using nuclear weapons against Iran. The US “publicly cited Iran as the culprit”.   Then Disstrack was used again in December 2015, in the attack that brought Sony to its knees.  These aren’t gangs using cybercrime for monetary gain. These are the equivalent of acts of war, given the level of damage done.

Fast forward to late 2016. Two major attacks happened in Saudi: November 17 taking out systems at the airport and other Saudi government agencies, and then again on November 29. Then, on January 23 there was another attack. The malware used was almost identical to the original Shamoon, aka Disstrack.  Except there were a few key enhancements.  According to Andrew Plato, CEO of Anitian Enterprise Security

 “What is really worrisome about this is it’s just outright destructive. It isn’t really trying to steal anything. It’s the closest things we’re going to get to a cyber bomb”.

The new version, dubbed Shamoon 2, spread through the local network using legitimate counts belonging to users and administrators, with complex passwords likely obtained from an earlier attack. Remember what I said about persistence?  This new version, however went on to attack VDIs, or Virtual Desktops, which previously could have offered some protection because of their ability to load snapshots of systems that were wiped. Now Shamoon had migrated from just Windows-based systems to Linux in the attacks on VDIs.

cyberwar1-1024x482

Now, I don’t want to be alarmist and spread FUD everywhere. Yes, this is serious and destructive. Like Stuxnet, it broke things. And that’s the differentiator. So far, the line hasn’t been crossed where breaking things was deliberately done to harm people. Because as Archer would say: You want cyberwar? Because that’s how you get cyberwar.

While the expectation is that Iran is once again behind the attacks, Symantec has revealed there are multiple parties involved. More than one entity, so collaboration and cooperation.  The report is that an entity known as Greenbug may have assisted in getting the credentials needed for access.  Palo Alto reported on a campaign known as Magic Hound which targeted energy, technology and government with ties or locations in Saudi.  There were links between Magic Hound and two other actors with Iranian ties: Charming Kitten and Rocket Kitten. Finally, putting all this together was the group Timberworm or Cobalt Gypsy.  Per Symantec, Timberworm was behind the January 23 attacks.

Here’s the play by play. First, Timberworm used spear phishing emails with weaponized documents (we warned you about those Office Macros!) to gain initial access into the network. Once there, they used custom malware, along with leveraging existing sysadmin tools to avoid detection, and help them achieve persistent remote access. Quick FYI: custom malware is a hallmark of major organized cybercrime groups or nation state attacks because it costs a lot of time and money to craft, and the stakes are going to be very high.

Apparently Greenbug and Timberworm have been active, penetrating organizations beyond Saudi. Note that Shamoon, however, was only used against the Saudi target. Timberworm is a large operation, as is Greenbug, with targets in a range of areas. We know who they are now, what they can do, and that they have a shared interest. What we don’t know: the endgame. I’m waiting for that other shoe to drop.

http://www.zerohedge.com/news/2016-12-01/another-false-flag-destructive-iranian-hackers-allegedly-wreak-havoc-saudi-computer-

http://www.securityweek.com/shamoon-2-variant-targets-virtualization-products

http://www.securityweek.com/multiple-groups-cooperated-shamoon-attacks-Symantec

http://www.archersecuritygroup.com/second-wave-bomb-malware-hits-saudi-arabia/

Ransomware Updates

We’ve got some new stuff out there. First, for those who torrent, be careful. If you torrent on a Mac, be very careful.  For the second time, ransomware has been designed for the Mac OS.In this case, “Patcher” is poor quality, shoddy code, to the extent that if the victim pays the ransom, they don’t get their files back because that code doesn’t work. It’s getting dropped via fake Adobe Premier Pro and Microsoft Office for Mac.

Second, if Google is telling you “Hoefler test not found”, don’t think you need to install that font. It’s a ploy on certain compromised websites to drop Spora ransomware. And very few AV or anti-malware programs can detect this one.

spora.JPG But, if you play it safe and do as Google says, click Discard and don’t download.  You’ll avoid ransomware.

If you want to know more, I’ve got a Ransomware page.

And saved the best for last. This amazing map from F-Secure shows the timeline of ransomware.  You can see the explosion that took place in 2016.

ransomware-tube-map

https://www.f-secure.com/documents/996508/1030743/cyber-security-report-2017

 

Banking on Insecurity

They came for the money, they stayed for the data. There is far more at stake in financial services than dollars and sense. The past twelve months have shown how far attackers are willing and able to go; banks are known for their conservative pace in adopting new strategies, and attackers are literally banking on it.

As the saying goes, “In God we trust”. In banks, maybe not so much.  According to a recent report by Capgemini, one in five bank execs are “highly confident” in their ability to detect a breach, never mind defend themselves against it.  Yet “83% of consumers believe their banks are secure from cyber attack”.  One in four banks report they’ve been attacked, but only 3% of consumers believe their bank has suffered a breach. Never mind the money. How about the data? Survey shows that 71% of banks don’t have a solid security strategy in place, nor do they have adequate data privacy practices. The numbers are not good. Only 40% of banking and insurance companies have automated security intelligence capabilities for proactive threat detection

After following the trail on the SWIFT bank heists last year, I’ve paid close attention to banking malware, threat actors, and points of failure. What worries me is what’s coming as digital payments become the norm, and digital identities take hold in developing nations who lack the infrastructure or regulation to secure or enforce. Given what we already know, what does this recent history of attacks tell us?

Polish Banks
The recent series of targeted malware attacks against Polish banks was identified in January this year, but attackers went after the data, not money. After noticing unusual network activity, like traffic to “exotic” locations and encrypted executables that nobody knew of, and unauthorised files on key machines in the network, several commercial banks confirmed malware infections. Investigations revealed infection stemmed from a tampered JS file from the webserver of the Polish financial sector regulatory body.  This was actually part of a wider campaign that has gone after financial institutions in over 30 countries.  According to researchers from both BAE Systems and Symantec, the malware used in Poland can be linked to similar attacks around the globe, and there are marked similarities to tools used by the cybercrime group Lazarus, although no confirmation has been made.  Targets were led to compromised sites of interest to them, watering holes, which were malicious sites that injected code and directed the targets to a customized exploit kit.  This kit contained exploits against known vulnerabilities in Flash Player and Silverlight. What’s interesting is that the exploits were only activated for certain visitors: those with IP addresses from specific ranges. Per Symantec, “The IP addresses belong to 104 different organizations located in 31 different countries … The vast majority of these organizations are banks, with a small number of telecoms and internet firms on the list.” 15 of these are from the US.  The infection downloaded enables recon on the compromised system. Again, this tool is similar to those used in past by the Lazarus group. Now every major security group has published their opinions and analysis on what was originally all but overlooked as some malware that spread from the regulatory body’s server.

Fileless Malware Attacks
In January of this year, there were reports around the globe of attacks on banks using fileless malware. The malware resided solely in the memory of compromised systems.  This is not signature based malware that can be referenced and detected. According to Kaspersky, 140 enterprises in 40 countries have been hit. And forensics cannot help us:

“ memory forensics is becoming critical to the analysis of malware and its functions. In these particular incidents, the attackers used every conceivable anti-forensic technique; demonstrating how no malware files are needed for the successful exfiltration of data from a network, and how the use of legitimate and open source utilities makes attribution almost impossible.” 

But the infections are hard to identify so that number could well be more.  Further complicating things is the use of legitimate and widely used sysadmin and security tools  like PowerShell, Metasploit and Mimikatz for malware injection. In a range of incidents, the common denominator seems to be embedding PowerShell in the registry to download Meterpreter. From there, the attack is carried out using the native Windows utilities and sysadmin tools. Per Kaspersky:

fileless1fileless2

The new fileless malware hitting banks is Duqu 2.0, which Kaspersky found on it corporate network in 2014, but only after it went undetected for 6 months because it lives almost completely in the memory of the computers. Duqu 2.0 is derived from Stuxnet. The malware renames itself when an infected computer is rebooted so digital forensics has a tough time finding traces. The calling card seems to be the unusual embedding of PowerShell into the registry to download Meterpreter. Duqu 2.0 is derived from Stuxnet. Reports aren’t saying how the malware spreads.

TESCO Bank Attack
In November 2016, Tesco Bank, a British retail bank chain with 7 million customers, warned its customers to watch for suspicious money withdrawals. Unfortunately, when customers who noticed money was missing from their accounts reached out to the bank, many could not get through. Approximately 20,000 accounts were hit. Tesco briefly halted online transactions in response. The attack seemed to stem from a “systemic failure of security around Tesco’s core database”. Recommendations include having controls in place to alert on changes to key files and configurations. As well, file monitoring integrity and Configuration Management Security ensure that if and when changes are made, they are valid and validated.

Take the Money and Run:  COBALT, ATMs and ‘Jackpotting’
There was a distinct rise in ATM attacks over 2016.  The latest siege, Cobalt, covers a wide swath across the UK, Spain, Russia, Romania, the Netherlands, much of Eastern Europe and Malaysia.  According to Group IB researchers, a large number of machines are attacked at once, and Cobalt appears to be linked to cybercrime syndicate Buhtrap.  The malware used causes infected machines to spit out cash in an attacks known as “jackpotting”.  Noteworthy is how this is being described as “the new model of organized crime”.  The FBI issued warnings to US banks following those ATM heists, taking into account the attacks in Taiwan and Thailand, when thieves grabbed over 260,000 pounds from Thailand’s Government savings bank and $2.5 million from Taiwan. The world’s two largest ATM manufacturers, NCR and Diebold Nixdorf, worked to manage the threat.

Lloyd’s Bank Hit by DDoS Attack
In January the venerable Lloyd’s Bank of London was struck by a DDoS attack that lasted two days.  Attackers tried to crash the Lloyd’s site, causing issues for customers and impacting some access to online banking.  The bank did not lose money, nor data, nor was the impact significant.  Law enforcement is investigating.

Attacks on Banks in the SWIFT System
Banks rely on messenger systems to conduct transfers back and forth. In 2016, a series of targeted attacks on banks in the trusted SWIFT messenger system came to light after a massive heist on the Bank of Bangladesh. Apparently the attacks are evolving, and SWIFT has told member bank, in an undisclosed letter from Nov. 2, that “attacks on its systems have only become more sophisticated in their strategies”.  “The threat is very persistent, adaptive and sophisticated – and it is here to stay”.  This is despite the work by regulators globally to toughen bank security measures. And the word is that “a fifth of them are hitting paydirt for the attackers”, per Stephen Gilderdale, head of SWIFT’s Customer Security Programme. Now the hackers exploit tech support software to gain access. Then send victims phony payment instructions via SWIFT network.  SWIFT emphasizes that all those attacks detected “exploited SWIFT interfaces used by its customers” but that the SWIFT communications network itself was not impacted. In light of this, warnings are being issued to small businesses to realize the threat to them is real.  Scams have become more sophisticated and will continue to evolve. 

Sources:

https://badcyber.com/several-polish-banks-hacked-information-stolen-by-unknown-attackers/
https://securelist.com/blog/research/77403/fileless-attacks-against-enterprise-networks/
https://www.symantec.com/connect/blogs/attackers-target-dozens-global-banks-new-malware-0
https://baesystemsai.blogspot.sk/2017/02/lazarus-watering-hole-attacks.html   https://threatpost.com/fileless-memory-based-malware-plagues-140-banks-enterprises/123652/
http://www.welivesecurity.com/2017/02/16/demystifying-targeted-malware-used-polish-banks/?utm_source=organic%20twitter&utm_medium=news&utm_campaign=WLS   http://economictimes.indiatimes.com/industry/banking/finance/banking/indian-banks-are-waking-up-to-a-new-kind-of-cyber-attack/articleshow/56575808.cms
https://www.f-secure.com/documents/996508/1030743/cyber-security-report-2017

My Approach to Threat Intel

In my role at work as a Threat Intel analyst, I track developments using various media feeds, and put together a succinct daily report of several key items that are pertinent to our clients and business lines.  Of course, I share my findings on Twitter and LinkedIn because that’s how the security community flourishes: collaboration. And to say I love what I do would be an understatement.

I don’t pretend to be an expert at what I do, nor will I say I have the definitive definition of what Threat Intel is. There is so much information to capture and analyze, and the learning is continuous. For me, my love of threat intel is in the hunt: looking for trends, patterns, new developments, things that reappear.  If you seek, you will find. There are many ways to search, and I am always trying to learn from people who have been doing this longer. It’s like fine-tuning a guitar, so I’ll always be looking at how to improve what I do.

I have go-to sources I read regularly, people online I follow specifically. My twitter feed is huge and categorized. But if I want to know something right away, it’s usually on there. I also have other sources to check in with directly. I collate information on malware, Advanced Persistent Threats (my most favourite things), specialized systems and their unique vulnerabilities.  This has helped me develop a baseline understanding over the time I’ve been doing this, so that I can understand who the players are when it comes to exploit kits, ransomware or DDoS.  And I try to make sure I know who the experts are, so that when they find something I am paying attention. That’s the head’s up.

When I’ve talked on Blue Teaming with my awesome pal, Haydn Johnson, we refer to the importance of knowing your baseline, watching patterns, so that you can identify anomalies. Those are your threats. That is your head’s up.  I find the same thing here as I track tweets, stories, advisories, reports and blogs.  I look for evolutions in how malware is delivered, so changes in exploit kits, or for kits to disappear from site. That means those kits are going to reappear with a new twist that our standard levels of detection and protection may not recognize, so attackers can access systems. Or, it could mean a larger scale attack, like Carbanak, when a massive crime gang operates on a global level and banks get taken for $1 billion. I play a lot of “what if” because I find I need to think beyond the normal realm to expect the unexpected. After all, the attackers are going where we aren’t looking.

In the weeks to come, I will be trying to bring in more information to widen my search. I’m researching all I can on what experts think best defines Threat Intel and Hunting. Because to really capture what’s out there, we need to broaden our scope.  I want to be looking ahead of the curve in this chase, anticipating their next move based on the wealth of information we have at hand, and factoring in what we know about human behavior. Next gen tech has spawned next gen threats, and as always, the attackers are ahead of us. And here is the thrill of the hunt.

Sector 2016

sector

This October marked the 10th anniversary of Toronto’s main security conference, Sector. I had the pleasure and privilege of being a speak, as well as working with a terrific team of volunteers. It was thrilling to be part of this event, plugged right in, to welcome people to our city and then to deliver a talk I had really wanted to give.

There was fanfare. Edward Snowden – yes, for real- was video conferenced in as the keynote speaker Day 1 and he did not disappoint. He has put his time away to good use, becoming expert on matters of privacy and rights. There was a second terrific keynote panel on Day 2 by a group of very successful and talented women about their experiences and insights on careers in InfoSec. The selection of talks and speakers was truly impressive, featuring leading experts and exciting new voices.

Here is my presentation, that started from a story on the Defensive Security podcast back in March. What caught my attention was how a bank heist in Bangladesh for a billion dollars was bungled because of a spelling error, and how far things almost went. Bank heists make great stories.  This year, we’ve got some really good stories to tell courtesy of a trusted network known as SWIFT, and some banks that believed they were inherently protected by virtue of being connected – except they weren’t. Hundreds of millions of dollars have revealed some ugly truths and dangerous assumptions.  In this security fairy tale we’ll talk about scary godmothers, big bad wolves, fire breathing dragons and what’s inherently wrong with the banking system. Because the emperors have no clothes on. Click on it to go to the site.

sectorslide

The Future of Ransomware

ransom

Ransomware is like like a nasty game of tag: you can try to avoid it but once you’re hit, you’re out. For all we know about doing defence right, following the best practices advocated by NIST and SANS, this particularly malevolent threat has been on an upward trajectory out of the gate since 2016, after trending through 2015.  It’s gone way beyond just phishing for targets and locking down individual files.  Current strains are evasive: like tag, they figure out what anti-virus and security is running on the target system that might detect it and stay hidden. They now go after websites. They lock down entire servers. And they don’t care who the victims are – not even hospitals.

Samsam-ransomware-attack-chain-768x391

If you’ve been reading along with me on Twitter, or happen to be up at 2:00 a.m. like I am, you know that ransomware is what keeps me up at night. Along with some other brilliant minds in our security community who are dedicated to tracking and shutting down this ever-growing threat. These guys really know what they’re doing. Countless hours of research, investigation and analysis have produced this paper:  Ransomware: Past, Present, and Future.   There are definitive pieces that give the lay of the land and map out the course ahead. That is what this piece does. Sincere appreciation for the efforts of  @da_667 @munin @ImmortanJo3 @wvualphasoldier (and others) who put this together. They understand just how widespread the risk is, and time is not a luxury we have. This is essential reading for anyone in tech, security, business, critical infrastructure. Essentially, anyone who needs to safeguard the data and networks their daily business relies on.

From the Talos blog: A fictional Adversary’s workflow of compromise and takeover

dadiagram

Right now, here is what I would advise anyone.  Back you stuff up, frequently, and separately from the network.  Check your patch management situation. Where are your exposures?  How are you handling security awareness, especially around phishing? Do you monitor your systems regularly, so that you have a baseline to compare events against?

And finally, take the time now and please read this: Ransomware: Past, Present and Future by Talos. Because the more people who know about ransomware and where it’s headed, the better we can all work together to secure things.

Thank you for stopping by!

My Layman’s Terms: The Java Deserialization Vulnerability in Current Ransomware

There has been a recent wave of ransomware attacks against hospitals, highly publicized and for good reason. Who the hell attacks hospitals with malicious code that locks up access to critical care systems, and puts our most vulnerable at further risk? Well, there’s more to this story than I can reveal here but I’ve been following the trend for months, and here’s what you need to know.

tweet ransom

FIRST: This was never about the hospitals. They weren’t the specific target. Law enforcement also relies on constant access to critical systems and they are being hit. But this goes so much wider, and we’re missing the bigger picture here. Therein lies the danger.   Samsa/Samsam has been a cash grab for the attackers, with no costs, no penalties. Don’t expect them to stop looking for more revenue streams to hit.

SECOND: This ransomware is not the same old ransomware. We can’t rely on our standard approaches to detect and defend against future attacks. This one goes after servers, so it can bring down entire networks, and doesn’t rely on the social engineering tactics to gain access.  It’s so bad US-CERT has issued this recent advisory.

I’ve laid out what’s been made available on just how this new strain of ransomware works. And I’ve done it in terms to help anybody take a closer look at the middleware running in their systems currently. Because a little knowledge could be dangerous thing used to our advantage this time.

tweetsamsa

WHAT: Extremely dangerous and wholly underated class of vulns

Attackers can gain complete remote control of an app server. Steal or corrupt data accessible from the server. Steal app code. Change the app. Use the server as launching oint for further attacks.

  • No working public exploits against apps til now
  • Remotely executable exploits against major middleware products
  • Powerful functionality that should not be exposed to untrusted users in the ability to hijack deserialization process.

IMPACT: Millions of app servers open to compromise

  • Not easily mitigated
  • Potential for millions of apps to be susceptible
  • Many enterprise apps vulnerable

AFFECTS: All apps that accept serialized Java objects

Remotely executable exploits against major middleware products:

  • WebSphere
  • WebLogic
  • JBoss
  • Jenkins
  • OpenNMS

HOW: Vulnerability is found in how many JAVA apps handle process of object deserialization.

Serialization is how programming languages transfer complex data structures over the network and between computers. Disassembly is the process of breaking an object down into a sequence of bits.

Deserialization is reassembly of those bits. (unserialization)

A Java object is broken down into series of bytes for easier transport.

Then is reassembled back at other end. Think the fly or tranporter

PROBLEM:  many applications that accept serialized objects do NOT validate or check UNTRUSTED input before deserialization or putting things back together. So yes, this is the perfect point to sneak the bad stuff in.

Attackers can INSERT malicious object into data stream and it can execute on the app server

Attack method:  special objects are serialized to cause the standard Java deserialization engine to instead run code the Attacker chooses.

Each of the 5 middleware applications listed above has a Java library called  “commons-collections.” This has a method that can lead to remote code execution when data is deserialized. Because no code should execute during this process.

NEEDS TO HAPPEN:

Enterprises must find all the places they use deserialized or untrusted data. Searching code alone will not be enough. Frameworks and libraries can also be exposed.

Need to harden it against the threat.

Removing commons collections from app servers will not be enough.   Other libraries can be affected.

Contrast Sec has a free tool for addressing issue.  Runtime Applicaton Self-Protection RASP.  Adds code to deserialization engine to prevent exploitation.

Sources:

Why the Java Deserialization Bug is a Big Deal Dark Reading by Jai Vijayan

What Do WebLogic, WebSphere, JBoss, Jenkins, OpenNMS, and Your Application Have in Common? This Vulnerability

Paypal is the latest victim of Java Deserialization Bugs in WebApps