Monday, July 25, 2016

Some thoughts about SWIFT banking incident

One of my recent gigs was result of  SWIFT banking incident in Bangladesh, Just to put thighs right I'm not expert on the malware analyses, my knowledge is related only to few online papers about incident  and chats with few law enforcement officers who were involved. My experience with SWIFT and banking is ancient one, last meddling was more than 10 years ago my last SWIFT system was installed on AIX machines.

Available resources in detail presents how attacked worked, high quality of attack code, results of investigations etc.. very much detailed from technical point. Since original incident  same attack was also done in fer other places.

What surprised me horribly was fact that SWIFT subsystem was implemented on windows machine,
I'm really curious why and how something so critical was put on windows and what was reasoning for that. Decision to use windows OS and SWIFT looks like disaster from any angle I can think of.  I really can't figure out what is a benefit of such solution. I really like to understand a) what was reason for SWIFT  to create windows version and b) what was the reason for bank to use windows version of SWIFT.

The scenario which comes to my mind is very much like one been explained to my why major forensic software vendors does not have UNIX versions. It is like our customers want us to use windows platform because this is the only platform they can efficiently use ... Basically it boils down to cutting expenses what is as we know from history a common cause of any horrible disaster
The reasoning chain behind this is very common among no-IT industries, where people do not understand their own business process, its critical dependency on IT and related risks. I'm quite confident we will see a lot of trouble of that type in future especially with "intelligent" devices and IOT (Internet of Things) expansion.

27.7.2016

So how can we rationally explain using swift on windows ? When you chat with people it comes out that the most common technical expertise is based on windows OS and MS based OS are majority in medium business and enterprise, so it means a huge base of people and low expenses. In the other hand UNIX based expertise is scare and not easy to find, sometimes it is even worse UNIX based systems and expertise are lot more expensive.  Using AIX machine for this purpose in completely windows based enterprise create small expensive island if you compare it only to production costs, but if you compare it to possible incident cost  it is quite cheap.  It is easy to prove, total breach was about 800 million dollars, while unrecoverd sum is about 80 millions, compared to 80 millions total costs of any AIX swift subsystem is invisible.
So we have here a rational chain of thought, decision based on available expertise, commonality of OS and not taking into account risks and its costs. Something to think about since this is not technical issues but organisational and managerial.

PS: I'm talking about AIX not because of IBM PR, but because my only SWIFT experience is with AIX. To be honest in my days I've seen few situations which were very deadly even for AIX/SWIFT combination but there was no breach.

1.8.2016
Very detailed report about bank interaction on Reuters site, worth of reading , The SWIFT hack How the New York Fed fumbled over the Bangladesh Bank cyber-heist

26.8.2016
I finally managed to chat about this story with friend and  coleagues. Different postions  and experiences from FIRST to banking regulatory.

1.9.2016
A new articles about new developments and new attacks.
When you think, for attacker it is best to have weak entry-point into system, like windows node, than to attack connection between bank and swift or bank IT. It is cheaper, since it is common point, in other hand each internal banking IT is a uniqe form of chaos ..

2.9.2016
From SANS newsbites

SWIFT Warns Member Banks of More Attacks
(August 31, 2016)
 
In February 2016, attackers stole US $81 million from Bangladesh Bank. In a letter to its clients earlier this week, global financial messaging system SWIFT disclosed that there have been more attacks, some successful, against member banks and urged them to adopt strong security measures.

Read more in:
-
 http://www.reuters.com

16.10.2016
Again From SANS newsbites, now everyone found that SWIFT systems are hackable

Odinaff Trojan Targets SWIFT System
(October 11, 2016)
 
Malware known as Odinaff is being used to target the SWIFT funds transfer system. Symantec says that roughly 100 organizations have been infected with Odinaff. The malware makes its way into systems by getting users to click on a malicious Microsoft office macro or password-protected RAR archive file.
Editor's Note

[William Hugh Murray]
Banks should use the indicators of compromise (IoCs) at https://www.symantec.com/security_response/writeup.jsp?docid=2016-083006-4847-99&tabid=2

Read more in:
- 
http://www.eweek.com: Odinaff Trojan Taking Aim at Financial Services
- http://www.theregister.co.uk: Second hacking group targets SWIFT-connected banks
- http://www.v3.co.uk: British banks targeted in new wave of Swift payments system attacks
- http://arstechnica.com: Emboldened by $1B Bangladesh hackers, new group targets SWIFT users
- http://www.computerworld.com: Second group of hackers found also targeting SWIFT users



19.12.2016 
There are more news about this story, looks like more incident happened and been "under-carpet-stored" interestingly not much fuss, Now it is normal that such "strong" organisation  fails with huge flop.. Banks, governments.. .. 
All this and yahoo breach story remembers me on old truth if you need action you must have a heap of dead bodies.

 7.3.2018
New stories about swift bank attacks about some banks in Russia. It is on the Sans news  bites. It show how a good investment pays off for a long time

Saturday, July 23, 2016

Probably a scam ...

A lot of connections coming trough this URL http://bit.ly/29ufKZW to this blog, It redirects to some fraud page somewhere far away ..  so be careful

Wednesday, July 6, 2016

Trying to get EnCase v8 ..

On the dreaded Guid portal  "Customer Community knowledge base" new version of EnCase is listed

Software


EnCase v8 - 8.01
EnCase v7 - 7.13
EnCase v6 - 6.19.7
Portable - 4.06.02
eDiscovery - 5.13
Endpoint- 5.13

But ... when you'll get only v7.12 when you try to download it ..

also customerservice@guidancesoftware.com. does not exist

KB is almost empty with one article about codemeter problems with v8


12. July 2016 today we just get announcement of v7.13 release, intead of 8 :)

It starts to be scary .. I suppose everyone is dead scared of version 8, Guid because of disaster with 7, customers because of getting something completely different and no one knows what from Guid.

It looks like a lost-lost situation

A than suddenly on completely unexpected account v8 arrived
looks like there is a v8.0.1 in wild

13. July 2016 v8 is not working with our edu dongles :)

It means no way to try and test trainign scenarios to see how it works. It is our practice for each release, since v7 get out into public, too see how trainign scenarios behave with new EnCase.
We always tried from EnCase Essentials,  Forensic 1 and Forensic 2 with some examples from Advanced Forensics training.  Encase essential is perfect since each dongle can open tdirden evidence file and customer can do test in environment.

2. August 2016

This morning we have announcement about v8 version of former EnCase Forensic I and EnCase Forensic II. Now there are new names:

EnCase Computer Forensics I is DF120-Foundations in Digital Forensics with EnCase Forensic.

EnCase Computer Forensics II is DF210-Building an Investigation with EnCase Forensic

What worries me my educational dongles are still not working with v8 and there is no official policy or announcement what and when this will happen. Looks like all training partners are in same trouble.

8.8.2016

looks like it is possible now to get a extension certificate for edu dongles, we will see if this will work
basically procedure is to contact customer support as ATP, send request for update with dongle id and order id.

20.8.2016

Still same, no version which can run on educational dongles

26.8.2016 

Finally organised and negotiated we will get a educational v8 dongles, our v7 dongles will get usual extension certificates for one year more, but we will get no NAS licence.
I feel tired it takes us as ATP almost a month to get v8.
As collateral looks like it is again possible to get demo versions of enterprise and related tools.
I am so tired and fed up of commercial forensic vendors and products
It is quite obvious why any attacker can wreck havoc ...

12.9.2016

Finally we get v8 dongles and it works !

20.10.2016 
There was announcement of new licencing server for integration of dongles or better to say NAS replacement with EnCase v7,14, instructions are extremely long and look complex, not a promising sign.

Thursday, June 30, 2016

Classroom and training preparation


Recently I've been involved in training abroad, mostly Middle east and Asia, some interesting figures shows up. There is procedure when we negotiate training in someoneelse classroom, we always send a minimal requirements to training partner. Basically we have to get confirmation if we can use classroom,  
Somehow this procedure failed, recently we have almost 75% of partners provided classrooms not satisfying minimal requirements.  Incredible combination of wrong configurations, wrong OS, heavy infected machines, broken hardware, wrong type of machines (MACs instead of PCs), power problems, everything bundled up,  you just  mention it. In all this situation there was a common line, we were negotiating with another company who was than facilitating things, in fact we were never in touch with technical staff responsible for classroom. A broken phone problem. Looks like this is unsolvable issue cultural thing, so how to prevent or mitigate such problems ?
Obvious solution, one where  to quit course and force partner to accept responsibility and pay penalities for mistakes will not work. So we are left with be smart, plan, prepare and adopt in your budget limits, This rules out "bringing the whole HW and SW class in one big pelican box" a wonderful solution but way to costly.
To be honest I love to have even a small pelican box with me, something reliable in wilderness, but 
we can only  afford approach where pelican box is small one , with minimal HW :) 

On what facts we can rely :

  1. classrooms are based on intel machines 
    • mixes of configurations, from dual core 32 bit machines up, 
    • minimum  USB 2.0 ports, bootable from USB
    • various localisations keyboard layouts / languages
    • most of machines does not have CD workable drive 
  1. networking is usually there (wifi or wired) but can't be relied on
  2. computers are in various states of OS and SW anarchy 

Strategy to win such classroom :

1) we will use existing classroom computers but in safe and reliable way,

  • boot each student PC into reliable and safe linux configuration and use appropriate virtual machine for student work
  • boot can be from USB or over network or from CD (rarely)

2) for clasroom server use trainer or trainer backup machine

  • again use virtual machine 
  •  boot from safe boot source, usb or network or from CD (rarely)

3) for networking use existing infrastructure or wifi over usb and 3G/4G hotspot on local mobile device, can be challenging

From this requirements you can easily draw a list of HW and SW which fits into one small pelican box ,  but a huge set of planning, testing and preparations, I'll cover that later

13.7.2016 List of devices and tools for such classrom



Items
Tablet / smartphone as local 3G/4G wifi hotspot, 8 conncetions
USB 3.0 stick with write block HW switch, 32 GB or more
USB 3.0 HUB with 4 ports and power
USB wifi module with windows & linux drivers
Labtops: 64 bit 4 core, 3 HD, 16 GB+, 4 USB 3.0 ports, GB etherent
USB 3.0 external disks and enclOsures
(with encRyprtIon)
CD with various linux distributions
USB CD rom / DVD
PCMCIA usb 3.0 card
USB 3.0/2.0 cables
USB 3.0/2.0 connectros / gender changers
Powersupplies for all devices and spares
Extension cords, power connetors etc
Tools, labels, stickers
Pelican box

Monday, June 13, 2016

Some thougths on research, education and valorisation for cybersecuirty

Dhaka,
Bangladesh,
11.6.2016

Vision on research, education and valorization for cybersecurity

There is a huge problem based on the introduction of modern deeply penetrating computer based technologies into society and into personal life of every individual. The term cybersecurity is just one small but crucial part of controlling this problem or better to say understanding it. We do not yet understand how and why these technologies will change our society, we don't even have reliable definitions of cyber and cyber-related issues. Even among professionals in the field we don't have complete understanding or a good intuition, I don't like to mention other involved but deeply ignorant parts of society, from general population to top decision makers. For some of this issues we can find parallels in the past, my deep concern is that we are not understanding this processes, maybe we are even using wrong methods to explore and analyze situation. Some of this events looks more like biological and medical than technical phenomena, more like great medieval plaques when we observe behavior and possible impacts on our society.
For this reasons I believe cybersecurity should be looked on as something essential for modern society, practically like a role medicine have achieved today with same organizational approach to the society highly trained and highly ethical professionals and widespread general knowledge with practice of hygiene, in this case recently developed cyberhygine. The analogy should be even wider, we should think about introducing biological and medical ideas and concepts into our approach to cybersecurity.
In lecturing and research, we should concentrate on the good general IT knowledge and technical perspective in various technologies, proving students with skills and ability for fast adopting new knowledge. We should widen the knowledge of students providing them with social, legal and historical perspective for events and technologies what is crucial, I believe, for understanding future events and trends. Such approach is currently painfully missing leaving students without knowledge about interaction among technology society, history and law.
For example, network security is crucial part of the cybersecurity but for most of the current networking curriculums, networking is presented as just set of standards and developments. Networking in a sense of security needs a holistic description of technology development and its impact on society. In that context parallels with US DOD approach in solving railway transport problems during civil war and solving communication problems introducing TCP/IP in cold war looks very similar with very same effects on society, there are commonalities even in morphing and developing a new types of crimes. Without such approach networking especially TCP/IP is just technical issues of set of protocols, not something opening new social development, a whole new wild west frontier. Providing such historical, social and legal context to teaching we enable students to grasp dynamics and get better understanding of current events and future developments. Applying this approach to cybersecurity we have to provide students not only with historical context, but with legislative and human context of crime and law also. As to get better understanding on human element of “cyber” I believe we can introduce reading fictions authors like Stanislav Lem, Isaac Asimov because of their intriguing insight.
For IT side we should provide students with IT skills in programming (especially defensive programming), scripting languages, theory of operating systems, networking, language theory, digital forensics, system and network administrations and security, big data handling, artificial intelligence and other relevant IT and science fields (especially practical mathematical knowledge).
Teaching should include practical work and theory but in a sense that student should be able to solve problems using scientific methods based on accepted theoretical knowledge, not just doing repetitive hands-on tasks or being frozen in theoretical framework. The key quality will be stress on analyzing problem, understanding it, finding solution and implementing it with evaluation of results, not just trying tools and raw computing power. Also we should stress the ethical approach and legal problems in solving complex real situations. By my observations we should also include more women in cybersecurity education, not only because of lack of women in cybersecurity but at least because of qualities of better group work in solving problems.
To achieve this goal there should be adequate technical resources (laboratories, classrooms, simulators with appropriate tools and equipment), cooperation with other academia, business, law enforcement locally and internationally. Practical work can be done in virtual and simulated environments but there should be student exposure to physical equipment and real working conditions, minimum of 10% of practical should be with real hardware. Academic research should provide framework and improve practicals while practical problems and solutions should be based on theoretically predicted scenarios or conquered real life events. There should be mandatory involvement of lecturing personnel in practicals and in supporting of CERT type organization with goal to keep practical skills up to date and understanding student community. I should suggest rotational approach with 25% of personnel in lecturing theory, 25% doing practical with students, 25% doing research, liaison and 25% in other activities. To keep with development practical should be modified or replaced yearly a good measure will be 30% changed per year, same for the theoretical part.


Wednesday, June 1, 2016

EnCase direct servlet preview

I've done a short ppt on how direct servlet is created and used in EnCase. Recently we often have such  questions so to simplify my work I've done small ppt which follows manual and put in on the Slideshare. It is easier to discuss with ppt than with going trough manual pages.

Monday, May 30, 2016

Managing digital forensic lab

For last few weeks I found myself in awkward situation, almost like echo from some previous jobs,
preparing materials for training titled "Managing digital forensic laboratory".  This is almost accidental event, a first run of that training since we announced it few years ago.
The story behind course is strange by itself too, It all started as result of a failure. There was one nice big project about setting up digital forensic lab, for dual purpose forensics and education. Huge effort was put into project, especially in preparation for managing lab materials and intro course for that. Since project didn't realise we did logical move, reused prepared materials, and squeezed into 3 day trainng expandable to 5 days.  Basic theory  is based on excellent "Building a Digital Forensic.Laboratory: Establishing and Managing a Successful Facility"book by Andrew Jones and Craig Valli. I've decided to add additional things based of recent development for datacenters and open source tools for compatibility and certification based on "Sarbanes-Oxley IT Compliance Using Open Source Tools, 2nd Edition" where we can show how to inexpensively build a managing and control infrastructure even on knoppix.  For lab case management I've decided to implement Foreman tool  and mention comparation with other case management tools like FTK lab.
To cover datacenter approach and introduce good practice in managing a lot of computing power, what is always missing  I've decided to use "Enterprise Data Center Design and Methodology"
By: Rob Snevely. There is a lot of other papers and web resources to mention and talk about ...